quorum/p2p/dial_test.go

706 lines
24 KiB
Go
Raw Normal View History

2015-07-06 17:54:22 -07:00
// Copyright 2015 The go-ethereum Authors
// This file is part of the go-ethereum library.
2015-07-06 17:54:22 -07:00
//
// The go-ethereum library is free software: you can redistribute it and/or modify
2015-07-06 17:54:22 -07:00
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// The go-ethereum library is distributed in the hope that it will be useful,
2015-07-06 17:54:22 -07:00
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
2015-07-06 17:54:22 -07:00
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
2015-07-06 17:54:22 -07:00
package p2p
import (
"encoding/binary"
"net"
"reflect"
"testing"
"time"
"github.com/davecgh/go-spew/spew"
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
"github.com/ethereum/go-ethereum/p2p/enode"
"github.com/ethereum/go-ethereum/p2p/enr"
"github.com/ethereum/go-ethereum/p2p/netutil"
)
func init() {
spew.Config.Indent = "\t"
}
type dialtest struct {
init *dialstate // state before and after the test.
rounds []round
}
type round struct {
peers []*Peer // current peer set
done []task // tasks that got done this round
new []task // the result must match this one
}
func runDialTest(t *testing.T, test dialtest) {
var (
vtime time.Time
running int
)
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
pm := func(ps []*Peer) map[enode.ID]*Peer {
m := make(map[enode.ID]*Peer)
for _, p := range ps {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
m[p.ID()] = p
}
return m
}
for i, round := range test.rounds {
for _, task := range round.done {
running--
if running < 0 {
panic("running task counter underflow")
}
test.init.taskDone(task, vtime)
}
new := test.init.newTasks(running, pm(round.peers), vtime)
if !sametasks(new, round.new) {
t.Errorf("round %d: new tasks mismatch:\ngot %v\nwant %v\nstate: %v\nrunning: %v\n",
i, spew.Sdump(new), spew.Sdump(round.new), spew.Sdump(test.init), spew.Sdump(running))
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
t.Log("tasks:", spew.Sdump(new))
// Time advances by 16 seconds on every round.
vtime = vtime.Add(16 * time.Second)
running += len(new)
}
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
type fakeTable []*enode.Node
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
func (t fakeTable) Self() *enode.Node { return new(enode.Node) }
func (t fakeTable) Close() {}
func (t fakeTable) LookupRandom() []*enode.Node { return nil }
func (t fakeTable) Resolve(*enode.Node) *enode.Node { return nil }
func (t fakeTable) ReadRandomNodes(buf []*enode.Node) int { return copy(buf, t) }
// This test checks that dynamic dials are launched from discovery results.
func TestDialStateDynDial(t *testing.T) {
runDialTest(t, dialtest{
init: newDialState(nil, nil, fakeTable{}, 5, nil),
rounds: []round{
// A discovery query is launched.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
new: []task{&discoverTask{}},
},
// Dynamic dials are launched when it completes.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&discoverTask{results: []*enode.Node{
newNode(uintID(2), nil), // this one is already connected and not dialed.
newNode(uintID(3), nil),
newNode(uintID(4), nil),
newNode(uintID(5), nil),
newNode(uintID(6), nil), // these are not tried because max dyn dials is 5
newNode(uintID(7), nil), // ...
}},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(3), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
},
// Some of the dials complete but no new ones are launched yet because
// the sum of active dial count and dynamic peer count is == maxDynDials.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(3), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(4), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(3), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
},
},
// No new dial tasks are launched in the this round because
// maxDynDials has been reached.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(3), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(4), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(5), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
new: []task{
&waitExpireTask{Duration: 14 * time.Second},
},
},
// In this round, the peer with id 2 drops off. The query
// results from last discovery lookup are reused.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(3), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(4), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(5), nil)}},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(6), nil)},
},
},
// More peers (3,4) drop off and dial for ID 6 completes.
// The last query result from the discovery lookup is reused
// and a new one is spawned because more candidates are needed.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(5), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(6), nil)},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(7), nil)},
&discoverTask{},
},
},
// Peer 7 is connected, but there still aren't enough dynamic peers
// (4 out of 5). However, a discovery is already running, so ensure
// no new is started.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(5), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(7), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(7), nil)},
},
},
// Finish the running node discovery with an empty set. A new lookup
// should be immediately requested.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(0), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(5), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(7), nil)}},
},
done: []task{
&discoverTask{},
},
new: []task{
&discoverTask{},
},
},
},
})
}
// Tests that bootnodes are dialed if no peers are connectd, but not otherwise.
func TestDialStateDynDialBootnode(t *testing.T) {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
bootnodes := []*enode.Node{
newNode(uintID(1), nil),
newNode(uintID(2), nil),
newNode(uintID(3), nil),
}
table := fakeTable{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
newNode(uintID(4), nil),
newNode(uintID(5), nil),
newNode(uintID(6), nil),
newNode(uintID(7), nil),
newNode(uintID(8), nil),
}
runDialTest(t, dialtest{
init: newDialState(nil, bootnodes, table, 5, nil),
rounds: []round{
// 2 dynamic dials attempted, bootnodes pending fallback interval
{
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
&discoverTask{},
},
},
// No dials succeed, bootnodes still pending fallback interval
{
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
},
// No dials succeed, bootnodes still pending fallback interval
{},
// No dials succeed, 2 dynamic dials attempted and 1 bootnode too as fallback interval was reached
{
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
},
// No dials succeed, 2nd bootnode is attempted
{
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(2), nil)},
},
},
// No dials succeed, 3rd bootnode is attempted
{
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(2), nil)},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(3), nil)},
},
},
// No dials succeed, 1st bootnode is attempted again, expired random nodes retried
{
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(3), nil)},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
},
// Random dial succeeds, no more bootnodes are attempted
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(4), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
},
},
},
})
}
func TestDialStateDynDialFromTable(t *testing.T) {
// This table always returns the same random nodes
// in the order given below.
table := fakeTable{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
newNode(uintID(1), nil),
newNode(uintID(2), nil),
newNode(uintID(3), nil),
newNode(uintID(4), nil),
newNode(uintID(5), nil),
newNode(uintID(6), nil),
newNode(uintID(7), nil),
newNode(uintID(8), nil),
}
runDialTest(t, dialtest{
init: newDialState(nil, nil, table, 10, nil),
rounds: []round{
// 5 out of 8 of the nodes returned by ReadRandomNodes are dialed.
{
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(2), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(3), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
&discoverTask{},
},
},
// Dialing nodes 1,2 succeeds. Dials from the lookup are launched.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(2), nil)},
&discoverTask{results: []*enode.Node{
newNode(uintID(10), nil),
newNode(uintID(11), nil),
newNode(uintID(12), nil),
}},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(10), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(11), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(12), nil)},
&discoverTask{},
},
},
// Dialing nodes 3,4,5 fails. The dials from the lookup succeed.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(10), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(11), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(12), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: dynDialedConn, dest: newNode(uintID(3), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(5), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(10), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(11), nil)},
&dialTask{flags: dynDialedConn, dest: newNode(uintID(12), nil)},
},
},
// Waiting for expiry. No waitExpireTask is launched because the
// discovery query is still running.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(10), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(11), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(12), nil)}},
},
},
// Nodes 3,4 are not tried again because only the first two
// returned random nodes (nodes 1,2) are tried and they're
// already connected.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(10), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(11), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(12), nil)}},
},
},
},
})
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
func newNode(id enode.ID, ip net.IP) *enode.Node {
var r enr.Record
if ip != nil {
r.Set(enr.IP(ip))
}
return enode.SignNull(&r, id)
}
// This test checks that candidates that do not match the netrestrict list are not dialed.
func TestDialStateNetRestrict(t *testing.T) {
// This table always returns the same random nodes
// in the order given below.
table := fakeTable{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
newNode(uintID(1), net.ParseIP("127.0.0.1")),
newNode(uintID(2), net.ParseIP("127.0.0.2")),
newNode(uintID(3), net.ParseIP("127.0.0.3")),
newNode(uintID(4), net.ParseIP("127.0.0.4")),
newNode(uintID(5), net.ParseIP("127.0.2.5")),
newNode(uintID(6), net.ParseIP("127.0.2.6")),
newNode(uintID(7), net.ParseIP("127.0.2.7")),
newNode(uintID(8), net.ParseIP("127.0.2.8")),
}
restrict := new(netutil.Netlist)
restrict.Add("127.0.2.0/24")
runDialTest(t, dialtest{
init: newDialState(nil, nil, table, 10, restrict),
rounds: []round{
{
new: []task{
&dialTask{flags: dynDialedConn, dest: table[4]},
&discoverTask{},
},
},
},
})
}
// This test checks that static dials are launched.
func TestDialStateStaticDial(t *testing.T) {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
wantStatic := []*enode.Node{
newNode(uintID(1), nil),
newNode(uintID(2), nil),
newNode(uintID(3), nil),
newNode(uintID(4), nil),
newNode(uintID(5), nil),
}
runDialTest(t, dialtest{
init: newDialState(wantStatic, nil, fakeTable{}, 0, nil),
rounds: []round{
// Static dials are launched for the nodes that
// aren't yet connected.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(3), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(5), nil)},
},
},
// No new tasks are launched in this round because all static
// nodes are either connected or still being dialed.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(3), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(3), nil)},
},
},
// No new dial tasks are launched because all static
// nodes are now connected.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(3), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(4), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(5), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(4), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(5), nil)},
},
new: []task{
&waitExpireTask{Duration: 14 * time.Second},
},
},
// Wait a round for dial history to expire, no new tasks should spawn.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(3), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(4), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(5), nil)}},
},
},
// If a static node is dropped, it should be immediately redialed,
// irrespective whether it was originally static or dynamic.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(3), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(5), nil)}},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(2), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(4), nil)},
},
},
},
})
}
// This test checks that static peers will be redialed immediately if they were re-added to a static list.
func TestDialStaticAfterReset(t *testing.T) {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
wantStatic := []*enode.Node{
newNode(uintID(1), nil),
newNode(uintID(2), nil),
}
rounds := []round{
// Static dials are launched for the nodes that aren't yet connected.
{
peers: nil,
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(2), nil)},
},
},
// No new dial tasks, all peers are connected.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(2), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(2), nil)},
},
new: []task{
&waitExpireTask{Duration: 30 * time.Second},
},
},
}
dTest := dialtest{
init: newDialState(wantStatic, nil, fakeTable{}, 0, nil),
rounds: rounds,
}
runDialTest(t, dTest)
for _, n := range wantStatic {
dTest.init.removeStatic(n)
dTest.init.addStatic(n)
}
// without removing peers they will be considered recently dialed
runDialTest(t, dTest)
}
// This test checks that past dials are not retried for some time.
func TestDialStateCache(t *testing.T) {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
wantStatic := []*enode.Node{
newNode(uintID(1), nil),
newNode(uintID(2), nil),
newNode(uintID(3), nil),
}
runDialTest(t, dialtest{
init: newDialState(wantStatic, nil, fakeTable{}, 0, nil),
rounds: []round{
// Static dials are launched for the nodes that
// aren't yet connected.
{
peers: nil,
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(2), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(3), nil)},
},
},
// No new tasks are launched in this round because all static
// nodes are either connected or still being dialed.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: staticDialedConn, node: newNode(uintID(2), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(1), nil)},
&dialTask{flags: staticDialedConn, dest: newNode(uintID(2), nil)},
},
},
// A salvage task is launched to wait for node 3's history
// entry to expire.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
done: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(3), nil)},
},
new: []task{
&waitExpireTask{Duration: 14 * time.Second},
},
},
// Still waiting for node 3's entry to expire in the cache.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
},
// The cache entry for node 3 has expired and is retried.
{
peers: []*Peer{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(1), nil)}},
{rw: &conn{flags: dynDialedConn, node: newNode(uintID(2), nil)}},
},
new: []task{
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
&dialTask{flags: staticDialedConn, dest: newNode(uintID(3), nil)},
},
},
},
})
}
func TestDialResolve(t *testing.T) {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
resolved := newNode(uintID(1), net.IP{127, 0, 55, 234})
table := &resolveMock{answer: resolved}
state := newDialState(nil, nil, table, 0, nil)
// Check that the task is generated with an incomplete ID.
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
dest := newNode(uintID(1), nil)
state.addStatic(dest)
tasks := state.newTasks(0, nil, time.Time{})
if !reflect.DeepEqual(tasks, []task{&dialTask{flags: staticDialedConn, dest: dest}}) {
t.Fatalf("expected dial task, got %#v", tasks)
}
// Now run the task, it should resolve the ID once.
config := Config{Dialer: TCPDialer{&net.Dialer{Deadline: time.Now().Add(-5 * time.Minute)}}}
srv := &Server{ntab: table, Config: config}
tasks[0].Do(srv)
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
if !reflect.DeepEqual(table.resolveCalls, []*enode.Node{dest}) {
t.Fatalf("wrong resolve calls, got %v", table.resolveCalls)
}
// Report it as done to the dialer, which should update the static node record.
state.taskDone(tasks[0], time.Now())
if state.static[uintID(1)].dest != resolved {
t.Fatalf("state.dest not updated")
}
}
// compares task lists but doesn't care about the order.
func sametasks(a, b []task) bool {
if len(a) != len(b) {
return false
}
next:
for _, ta := range a {
for _, tb := range b {
if reflect.DeepEqual(ta, tb) {
continue next
}
}
return false
}
return true
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
func uintID(i uint32) enode.ID {
var id enode.ID
binary.BigEndian.PutUint32(id[:], i)
return id
}
// implements discoverTable for TestDialResolve
type resolveMock struct {
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
resolveCalls []*enode.Node
answer *enode.Node
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
func (t *resolveMock) Resolve(n *enode.Node) *enode.Node {
t.resolveCalls = append(t.resolveCalls, n)
return t.answer
}
all: new p2p node representation (#17643) Package p2p/enode provides a generalized representation of p2p nodes which can contain arbitrary information in key/value pairs. It is also the new home for the node database. The "v4" identity scheme is also moved here from p2p/enr to remove the dependency on Ethereum crypto from that package. Record signature handling is changed significantly. The identity scheme registry is removed and acceptable schemes must be passed to any method that needs identity. This means records must now be validated explicitly after decoding. The enode API is designed to make signature handling easy and safe: most APIs around the codebase work with enode.Node, which is a wrapper around a valid record. Going from enr.Record to enode.Node requires a valid signature. * p2p/discover: port to p2p/enode This ports the discovery code to the new node representation in p2p/enode. The wire protocol is unchanged, this can be considered a refactoring change. The Kademlia table can now deal with nodes using an arbitrary identity scheme. This requires a few incompatible API changes: - Table.Lookup is not available anymore. It used to take a public key as argument because v4 protocol requires one. Its replacement is LookupRandom. - Table.Resolve takes *enode.Node instead of NodeID. This is also for v4 protocol compatibility because nodes cannot be looked up by ID alone. - Types Node and NodeID are gone. Further commits in the series will be fixes all over the the codebase to deal with those removals. * p2p: port to p2p/enode and discovery changes This adapts package p2p to the changes in p2p/discover. All uses of discover.Node and discover.NodeID are replaced by their equivalents from p2p/enode. New API is added to retrieve the enode.Node instance of a peer. The behavior of Server.Self with discovery disabled is improved. It now tries much harder to report a working IP address, falling back to 127.0.0.1 if no suitable address can be determined through other means. These changes were needed for tests of other packages later in the series. * p2p/simulations, p2p/testing: port to p2p/enode No surprises here, mostly replacements of discover.Node, discover.NodeID with their new equivalents. The 'interesting' API changes are: - testing.ProtocolSession tracks complete nodes, not just their IDs. - adapters.NodeConfig has a new method to create a complete node. These changes were needed to make swarm tests work. Note that the NodeID change makes the code incompatible with old simulation snapshots. * whisper/whisperv5, whisper/whisperv6: port to p2p/enode This port was easy because whisper uses []byte for node IDs and URL strings in the API. * eth: port to p2p/enode Again, easy to port because eth uses strings for node IDs and doesn't care about node information in any way. * les: port to p2p/enode Apart from replacing discover.NodeID with enode.ID, most changes are in the server pool code. It now deals with complete nodes instead of (Pubkey, IP, Port) triples. The database format is unchanged for now, but we should probably change it to use the node database later. * node: port to p2p/enode This change simply replaces discover.Node and discover.NodeID with their new equivalents. * swarm/network: port to p2p/enode Swarm has its own node address representation, BzzAddr, containing both an overlay address (the hash of a secp256k1 public key) and an underlay address (enode:// URL). There are no changes to the BzzAddr format in this commit, but certain operations such as creating a BzzAddr from a node ID are now impossible because node IDs aren't public keys anymore. Most swarm-related changes in the series remove uses of NewAddrFromNodeID, replacing it with NewAddr which takes a complete node as argument. ToOverlayAddr is removed because we can just use the node ID directly.
2018-09-24 15:59:00 -07:00
func (t *resolveMock) Self() *enode.Node { return new(enode.Node) }
func (t *resolveMock) Close() {}
func (t *resolveMock) LookupRandom() []*enode.Node { return nil }
func (t *resolveMock) ReadRandomNodes(buf []*enode.Node) int { return 0 }