Commit Graph

29 Commits

Author SHA1 Message Date
Leo 6c7ca964e4 Remove outdated TODO comments
- Metrics tracked in #11.
- Timeout and retransmits covered in #21.
- Dependency injection doesn't make sense at this scale.
- `-1` on `GenerateKeyPair` means "this doesn't make sense for Ed25519,
  please crash if anyone ever tried to generate RSA keys".

ghstack-source-id: 8951628351
Pull Request resolved: https://github.com/certusone/wormhole/pull/66
2020-10-28 22:47:12 +01:00
Leo 11c74dd692 bridge: refactor p2p logic into pkg/p2p
ghstack-source-id: 86417c1301
Pull Request resolved: https://github.com/certusone/wormhole/pull/65
2020-10-28 22:47:12 +01:00
Leo fd27570637 bridge: remove all supervisor.SignalHealthy calls
Supervisor does not back off tasks that failed in a healthy state.

There are a couple places where we rely on supervisor for
application-level backoff, so we always want back-off. The distinction
is meant to enable runnables to implement their own specific back-off
logic, which we don't, so we can safely ignore it.

Fixes #37

ghstack-source-id: c756381b1b
Pull Request resolved: https://github.com/certusone/wormhole/pull/64
2020-10-28 22:47:12 +01:00
Leo 8be34ac657 bridge: refactor processor logic into pkg/processor
This increases modularity in preparation for integration testing
and addition of a CLI command hierarchy.

ghstack-source-id: 9c94373236
Pull Request resolved: https://github.com/certusone/wormhole/pull/63
2020-10-28 22:47:12 +01:00
Leo 509d9c3d73 bridge: use mlockall(2) to prevent pages from being swapped out
The extra capability is harmless and is, at worst, a DoS risk.

ghstack-source-id: d30b50dbca
Pull Request resolved: https://github.com/certusone/wormhole/pull/55
2020-10-22 16:51:56 +02:00
Leo 3b35ddc1ce bridge: propagate panics from runnables
Any error that can be recovered by restarting a runnable
can also be recovered from by restarting the entire process.

If we encounter a panic, it's safer to restart the process than
attempting to limp along by restarting the runnable.

We always assume that an external process manager will restart
our process if it crashes. We already rely on this behavior for
libp2p errors which we handle by terminating the process, since libp2p
maintains global state that we can't clear.

ghstack-source-id: 2f20ef764d
Pull Request resolved: https://github.com/certusone/wormhole/pull/45
2020-10-22 12:20:11 +02:00
Hendrik Hofstadt d537f976f9 bridge: setup pubsub before connecting to bootnodes; buffer observations 2020-08-31 16:11:09 +02:00
Leo ef2aab5998 Decouple lifecycle of processor and ethwatch
We now do an independent fetch of the guardian set.
2020-08-22 00:21:57 +02:00
Leo 72289be8ee Wire up Solana lockup watcher 2020-08-21 20:49:33 +02:00
Leo b663e2dc56 Generalize token lockup processor 2020-08-21 13:00:44 +02:00
Leo 10621625f0 Handle supervisor.RunGroup error 2020-08-21 13:00:44 +02:00
Leo 935411c036 Refactor all node key management into nodekeys.go 2020-08-21 13:00:44 +02:00
Leo d317164345 Rename ethwatch to ethlockup and run it in a runnable group with the watcher
This ensures that we re-fetch the guardian set when ethlockup dies.
2020-08-21 11:17:55 +02:00
Leo f1f2d0c8b8 Solana VAA submission stub 2020-08-20 21:48:58 +02:00
Leo d6ef9c932c bridge: listen to eth lockups and aggregate signatures from all nodes
Improved devnet setup to generate deterministic node and guardian keys.

Devnet setup routine that configures a dynamic guardian set on Ethereum.

Configurable number of nodes in Tiltfile.
2020-08-19 14:24:38 +02:00
Leo a1dcc36df9 Handle ipfslog.SetLogLevel error 2020-08-17 23:02:06 +02:00
Leo 760d5541b7 Move loadGuardianKey and ethLockupProcessor out of main.go 2020-08-17 19:36:17 +02:00
Leo 7903402fa6 Deterministic hashes for ETH lockups
We're missing a nonce for truly unique hashes - for now, two
identical transfers will only be executed once.
2020-08-17 19:29:25 +02:00
Leo bc3714fc73 Add nodeName to override hostname in gossip 2020-08-17 19:22:12 +02:00
Leo 955bcc93fe bridge: devnet guardian key generation 2020-08-17 18:20:28 +02:00
Leo e548bbbee3 devnet: add send-lockups.js truffle script to generate test lockups 2020-08-17 16:31:48 +02:00
Leo c4d53247d3 bridge: fix p2p routine restart
Turns out, libp2p cannot be cleaned up so till this is fixed,
we have to exit if p2p dies.

Expose the root context to p2p and cancel it.
2020-08-17 14:55:51 +02:00
Leo 090d0aca84 bridge: split up guardiand/main.go 2020-08-17 12:29:52 +02:00
Leo bdf164a5b5 bridge: signal readiness 2020-08-17 12:25:16 +02:00
Leo 24d38b12ea bridge: default to 15 blocks ETH lockup 2020-08-17 12:14:57 +02:00
Leo 2744c1df25 bridge: heartbeat, eth watcher service 2020-08-16 17:05:58 +02:00
Leo 28fef7efca bridge: make sure the process crashes if we can't connect to any bootstrap peers
Supervisor rescheduling doesn't do the trick since the p2p socket doesn't clean up reliably.
2020-08-16 15:02:11 +02:00
Leo a7d17cf5aa Add persistent keys and bootstrap peer
This connects all nodes to a full mesh.
2020-08-04 12:15:51 +02:00
Leo dd8b2ade81 libp2p stub that can find peers on a private network
Test Plan:

    go run github.com/certusone/wormhole/bridge/cmd/guardiand -network wormhole-dev -port 9000 &
    go run github.com/certusone/wormhole/bridge/cmd/guardiand -network wormhole-dev -bootstrap /ip4/127.0.0.1/udp/9000/quic/p2p/<id> -port 9001
2020-08-04 12:15:51 +02:00