* Update Connection debug impl missed in previous PRs
* Add an initial_cached_addrs argument to Connection::new()
* Stop sending version and remote IP peers directly to the address book
* Update zebra-network/src/peer/connection.rs
Co-authored-by: teor <teor@riseup.net>
---------
Co-authored-by: Marek <mail@marek.onl>
* Avoid a race condition in the 25.3.0 state upgrade
* Look for the most recent sprout tree height if needed
* Get the latest history tree not the tip height history tree
* Discard keys provided by low level database method
* Remove extra ;
* Provide key types & rustfmt
* Fix weird closure type syntax
* Update zebra-state/src/service/finalized_state/disk_format/upgrade/fix_tree_key_type.rs
Co-authored-by: Marek <mail@marek.onl>
---------
Co-authored-by: Marek <mail@marek.onl>
* Skip external branches in ci-unit-tests-docker.yml
* Patch external repositories in ci-unit-tests-docker.patch.yml
* Revert ci-unit-tests-docker.patch.yml
* Add a separate external PR patch in ci-unit-tests-docker.patch-external.yml
* Document ci-unit-tests-docker.patch.yml
* Document ci-unit-tests-docker.yml
* Skip dependabot config validation for external PRs
* Trivial Rust code modification to make Rust workflows run
* Use correct method for checking source branch refs
* Fix the check condition
* Simplify job structure
* Use correct skip condition
* Make the start docs visible so we can link to them
* Skip integration tests on external PRs
* Allow external PRs to pass CD Docker tests
* Allow external PRs to pass Docs build and deploy
* Updates ADDR_RESPONSE_LIMIT_DENOMINATOR to 4
* Moves logic getting a fraction of Zebra's peers to a method in the address book
* Adds and uses CachedPeerAddrs struct in inbound service
* moves and documents constant
* fixes test
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* updates docs
* renames sanitized_window method
* renames CachedPeerAddrs to CachedPeerAddrResponse
* updates test
* moves try_refresh to per request
* Make unused sanitization method pub(crate)
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* moves CachedPeerAddrResponse to a module
* updates unit test
* fixes unit test
* removes unnecessary condition
* clears cached getaddr response if it can't refresh for over a minute after the refresh time
* tests that inbound service gives out the same addresses for every Peers request before the refresh interval
* Applies suggestion from code review
* fixes doc link
* renames constant
* Fix docs on new constant
* applies suggestion from code review
* uses longer cache expiry time
* Adds code comments
---------
Co-authored-by: teor <teor@riseup.net>
* Reformat a method to make related code clearer
* Limit how many addresses we take from each peer response
* Only put extra addresses back in the cache if there are actually any extras
* Refactor address caching into a single method
* Use the same method for updating the peer cache and taking peers from it
* Choose response peers at random, and partially randomly drop excess peers
* Update docs for this fix
* Fix an index bug and clarify a comment
* Add a proptest for cache updates
* Refactor return type of poll_discover()
* Simplify poll_ready() by removing preselected peers
* Fix peer set readiness check
* Pass task context correctly to background tasks
* Make poll_discover() return Pending
* Make poll_inventory() return Pending
* Make poll_unready() return Poll::Pending
* Simplify with futures::ready!() and ?
* When there are no peers, wake on newly ready peers, or new peers, in that order
* Preserve the original waker when there are no unready peers
* Fix polling docs and remove unnecessary code
* Make sure we're ignoring Poll::Pending not Result
* Make panic checking method names clearer
* Fix connection Client task wakeups and error handling
* Cleanup connection panic handling and add wakeup docs
* Fix connection client task wakeups to prevent hangs
* Simplify error and pending handling
* Clarify inventory set behaviour
* Define peer set poll_* methods so they return Ok if they do something
* Clarify documentation
Co-authored-by: Arya <aryasolhi@gmail.com>
* Fix test that depended on preselected peers
* Check ready peers for errors before sending requests to them
* Fix a hanging test by not waiting for irrelevant actions
* Only remove cancel handles when they are required
* fix incorrect panic on termination setting
* Clarify method comments
Co-authored-by: Arya <aryasolhi@gmail.com>
---------
Co-authored-by: Arya <aryasolhi@gmail.com>
* add a basic RAM database for the scanner
* specify a crate version for zebra-chain dependency
* add a type for sapling keys
* rename everything to reflect sapling
* change some storage methods
* get started with the blockchain scanner poc
* rustfmt
* fix the tests
* Reads blocks from db
* Adds conversion functions
* scans blocks and counts transactions
* fix bug
* split into 2 tests
* add duplicated dependencies to deny.toml
* upgrade zebra-scanner version
* try removing ecc duplicated dependencies
* try fix deny.toml
* remove unintended paste from deny.toml
* remove duplicated code from the other test
* remove strict version of `zcash_primitives` crate
* change description
* remove feture
* remove tokio features
* change lib doc
* add more documentation
* change expect
* do not use default in compact block creation
* more docs
* add more checks to test
* remove zebra-consensus dependency
* move all deps to dev-deps
* change crate version
* rename crate to zebra-scan
* lock file
* add test for zecpages populated state
* scans all cached blocks for zecpages viewing key expecting Ok results.
* use test blocks
* fixes test
* fix expect messages
* Discard changes to Cargo.lock
* Discard changes to deny.toml
---------
Co-authored-by: arya2 <aryasolhi@gmail.com>
Co-authored-by: teor <teor@riseup.net>
* Stop using `SemanticallyVerifiedBlockWithTrees`
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* Doc that default treestate corresponds to genesis
* Doc the diff between default values of tree roots
---------
Co-authored-by: teor <teor@riseup.net>
* get started with the blockchain scanner poc
* rustfmt
* fix the tests
* Reads blocks from db
* Adds conversion functions
* scans blocks and counts transactions
* fix bug
* split into 2 tests
* add duplicated dependencies to deny.toml
* upgrade zebra-scanner version
* try removing ecc duplicated dependencies
* try fix deny.toml
* remove unintended paste from deny.toml
* remove duplicated code from the other test
* remove strict version of `zcash_primitives` crate
* change description
* remove feture
* remove tokio features
* change lib doc
* add more documentation
* change expect
* do not use default in compact block creation
* more docs
* add more checks to test
* remove zebra-consensus dependency
* move all deps to dev-deps
* change crate version
* rename crate to zebra-scan
* lock file
* ifix cargo.lock
* remove internal dev dependencies versions
Co-authored-by: teor <teor@riseup.net>
* fix docs url
Co-authored-by: teor <teor@riseup.net>
* fix expect messages
Co-authored-by: teor <teor@riseup.net>
* remove duplicated in deny.toml
Co-authored-by: teor <teor@riseup.net>
* add a comment about moving code to production
---------
Co-authored-by: arya2 <aryasolhi@gmail.com>
Co-authored-by: teor <teor@riseup.net>
* Fix a potential hang accessing the address book directly
* Remove unused connection shutdown MetaAddr arguments
* Add an UpdateConnected MetaAddrChange, that sends initial connection info
* Fix some tests
* Fix a panic with a zero channel size
* Update PR template checklists
* Add a testing section
* Move close issues to the top section
* Change checklist phrasing so the answer is yes when the PR is complete
* Explain how to handle irrelevant checkboxes
* Another yes answer is good
* Fix indent
* Ask if PRs can be split
* Remove DockerHub tags for v1.x.y and edge
* Make the latest tag point to the production image, not experimental
* Remove docker image file change trigger from the OS CI workflow
* Disable latest tag by default
* Enable the latest tag for the production build only
* Restore `edge` tag, it is used in CI and it is CI only
* Add Docker tag removal to CHANGELOG.md
* Actually enforce job order
* Revert tag publishing changes, but keep extra docs and tag order changes
* Update the changelog
* Explain how to change branch protection rules
* Fix list formatting
* Add missing .
* Fix incorrect link
Co-authored-by: Pili Guerra <mpguerra@users.noreply.github.com>
---------
Co-authored-by: Pili Guerra <mpguerra@users.noreply.github.com>
* Refactor out try_to_sync_once()
* Add outer timeouts for obtaining and extending tips
* Refactor out request_genesis_once()
* Wrap genesis download once in a timeout
* Increase the genesis timeout to avoid denial of service from old nodes
* Add an outer timeout to mempool crawls
* Add an outer timeout to mempool download/verify
* Remove threaded mutex blocking from the inbound service
* Explain why inbound readiness never hangs
* Fix whitespace that cargo fmt doesn't
* Avoid hangs by always resetting the past lookahead limit flag
* Document block-specific and syncer-wide errors
* Update zebrad/src/components/sync.rs
Co-authored-by: Marek <mail@marek.onl>
* Use correct condition for log messages
Co-authored-by: Marek <mail@marek.onl>
* Keep lookahead reset metric
---------
Co-authored-by: Arya <aryasolhi@gmail.com>
Co-authored-by: Marek <mail@marek.onl>
* Simple replacements of doc.zebra.zfnd.org with docs.rs
* Manual fixes for specific main/internal/external docs
* Point developer docs to doc-internal.zebra.zfnd.org
* fastmod --glob '\!.git' -- doc.zebra.zfnd.org/zebrad docs.rs/zebrad/latest/zebrad
* Manually remove any remaining doc.zfnd.zebra.org links
* Remove the external docs job
* Add changelog entry and fix links
* Fix links that were broken before this PR
* Fixes bug where Zebra won't reconnect to peers after brief loss of network connectivity
* only dial on timercrawl when theres a new address or zero active outbound conns
Alfredo Garcia
teor
Arya
dependabot[bot]
Marek