This can happen under heavy load - no value in warning here.
Scenario this happens:
- New slot
- We get asked about what fork to build on
- We start building (create inherent gets called)
- We learn about a better fork (but Babe does not care as it already has a fork)
- 6 seconds passed - New slot
- Babe asks about what for to build on - we answer with the best block we learned about 6 seconds ago (slightly after we got asked the last time)
- We start building (on that old block)
- Milliseconds later we learn about a new block (the one from the slot we are actually in now)
- We kick the old leaf -> create inherent fails
Full discussion: https://github.com/paritytech/substrate/issues/12375
* Don't store available data on disputes
If there are lots of disputes, this leads to blowing up disk space on
validators. Rob luckily remembered that we do store the full
availability in participation.
The argument in the code does not make too much sense with the current
implementation, as no validator will ever request anything else from us,
than the one piece we are meant to posess.
* Fix warnings.
* Fix compile warnings
* Remove redundant field.
Co-authored-by: Vsevolod Stakhov <vsevolod.stakhov@parity.io>
* Don't import backing statements directly
into the dispute coordinator. This also gets rid of a redundant
signature check. Both should have some impact on backing performance.
In general this PR should make us scale better in the number of parachains.
Reasoning (aka why this is fine):
For the signature check: As mentioned, it is a redundant check. The
signature has already been checked at this point. This is even made
obvious by the used types. The smart constructor is not perfect as
discussed [here](https://github.com/paritytech/polkadot/issues/3455),
but is still a reasonable security.
For not importing to the dispute-coordinator: This should be good as the
dispute coordinator does scrape backing votes from chain. This suffices
in practice as a super majority of validators must have seen a backing
fork in order for a candidate to get included and only included
candidates pose a threat to our system. The import from chain is
preferable over direct import of backing votes for two reasons:
1. The import is batched, greatly improving import performance. All
backing votes for a candidate are imported with a single import.
And indeed we were able to see in metrics that importing votes
from chain is fast.
2. We do less work in general as not every candidate for which
statements are gossiped might actually make it on a chain. The
dispute coordinator as with the current implementation would still
import and keep those votes around for six sessions.
While redundancy is good for reliability in the event of bugs, this also
comes at a non negligible cost. The dispute-coordinator right now is the
subsystem with the highest load, despite the fact that it should not be
doing much during mormal operation and it is only getting worse
with more parachains as the load is a direct function of the number of statements.
We'll see on Versi how much of a performance improvement this PR
* Get rid of dead code.
* Dont send approval vote
* Make it pass CI
* Bring back tests for fixing them later.
* Explicit signature check.
* Resurrect approval-voting tests (not fixed yet)
* Send out approval votes in dispute-distribution.
Use BTreeMap for ordered dispute votes.
* Bring back an important warning.
* Fix approval voting tests.
* Don't send out dispute message on import + test
+ Some cleanup.
* Guide changes.
Note that the introduced complexity is actually redundant.
* WIP: guide changes.
* Finish guide changes about dispute-coordinator
conceputally. Requires more proof read still.
Also removed obsolete implementation details, where the code is better
suited as the source of truth.
* Finish guide changes for now.
* Remove own approval vote import logic.
* Implement logic for retrieving approval-votes
into approval-voting and approval-distribution subsystems.
* Update roadmap/implementers-guide/src/node/disputes/dispute-coordinator.md
Co-authored-by: asynchronous rob <rphmeier@gmail.com>
* Review feedback.
In particular: Add note about disputes of non included candidates.
* Incorporate Review Remarks
* Get rid of superfluous space.
* Tidy up import logic a bit.
Logical vote import is now separated, making the code more readable and
maintainable.
Also: Accept import if there is at least one invalid signer that has not
exceeded its spam slots, instead of requiring all of them to not exceed
their limits. This is more correct and a preparation for vote batching.
* We don't need/have empty imports.
* Fix tests and bugs.
* Remove error prone redundancy.
* Import approval votes on dispute initiated/concluded.
* Add test for approval vote import.
* Make guide checker happy (hopefully)
* Another sanity check + better logs.
* Reasoning about boundedness.
* Use `CandidateIndex` as opposed to `CoreIndex`.
* Remove redundant import.
* Review remarks.
* Add metric for calls to request signatures
* More review remarks.
* Add metric on imported approval votes.
* Include candidate hash in logs.
* More trace log
* Break cycle.
* Add some tracing.
* Cleanup allowed messages.
* fmt
* Tracing + timeout for get inherent data.
* Better error.
* Break cycle in all places.
* Clarified comment some more.
* Typo.
* Break cycle approval-distribution - approval-voting.
Co-authored-by: asynchronous rob <rphmeier@gmail.com>
We are awaiting on the oneshot anyways, so we have back pressure. By
using the unbounded channel make log messages like the following less
likely (due to higher priority):
2022-05-30 13:46:38
2022-05-30 11:46:38.565 WARN tokio-runtime-worker parachain::provisioner: failed to assemble or send inherent data err=CanceledBackedCandidates(Canceled)
* Double grandpa gossip duration.
* Make resend period slightly larger.
So it won't get triggered by additional grandpa delay.
* Bump other values as well.
* Don't change gossip duration on Polkadot.
(and Westend as it is meant to be a testbed for Polkadot)
* Don't wait for dispute coordinator
in backing and approval-voting - we are single threaded there, so this
is blocking everything.
* Add missing import.
* Don't warn on dropped receiver.
* First step in implementing https://github.com/paritytech/polkadot/issues/4386
This PR:
- Reduces MAX_UNSHARED_UPLOAD_TIME to 150ms
- Increases timeout on collation fetching to 1200ms
- Reduces limit on needed backing votes in the runtime
This PR does not yet reduce the number of needed backing votes on the
node as this can only be meaningfully enacted once the changed limit in
the runtime is live.
* Fix tests.
* Guide updates.
* Review remarks.
* Bump minimum required backing votes to 2 in runtime.
* Make sure node side code won't make runtime vomit.
* cargo +nightly fmt
* Versi chainspec based on Rococo.
* Fix.
* Fix compilation without features.
* Spelling
* Another typo.
* Support versi in native version
* 🤦
* Regenerated genesis for versi.
* Update cli/src/command.rs
Co-authored-by: Bastian Köcher <info@kchr.de>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
* Mostly notes.
* Better error messages.
* Introduce Fatal/NonFatal + drop back channel participation
- Fatal/NonFatal - in order to make it easier to use utility functions.
- We drop the back channel in dispute participation as it won't be
needed any more.
* Better error messages.
* Utility function for receiving `CandidateEvent`s.
* Ordering module typechecks.
* cargo fmt
* Prepare spam slots module.
* Implement SpamSlots mechanism.
* Implement queues.
* cargo fmt
* Participation.
* Participation taking shape.
* Finish participation.
* cargo fmt
* Cleanup.
* WIP: Cleanup + Integration.
* Make `RollingSessionWindow` initialized by default.
* Make approval voting typecheck.
* Get rid of lazy_static & fix approval voting tests
* Move `SessionWindowSize` to node primitives.
* Implement dispute coordinator initialization.
* cargo fmt
* Make queues return error instead of boolean.
* Initialized: WIP
* Introduce chain api for getting finalized block.
* Fix ordering to only prune candidates on finalized events.
* Pruning of old sessions in spam slots.
* New import logic.
* Make everything typecheck.
* Fix warnings.
* Get rid of obsolete dispute-participation.
* Fixes.
* Add back accidentelly deleted Cargo.lock
* Deliver disputes in an ordered fashion.
* Add module docs for errors
* Use type synonym.
* hidden docs.
* Fix overseer tests.
* Ordering provider taking `CandidateReceipt`.
... To be kicked on one next commit.
* Fix ordering to use relay_parent
as included block is not unique per candidate.
* Add comment in ordering.rs.
* Take care of duplicate entries in queues.
* Better spam slots.
* Review remarks + docs.
* Fix db tests.
* Participation tests.
* Also scrape votes on first leaf for good measure.
* Make tests typecheck.
* Spelling.
* Only participate in actual disputes, not on every import.
* Don't account backing votes to spam slots.
* Fix more tests.
* Don't participate if we don't have keys.
* Fix tests, typos and warnings.
* Fix merge error.
* Spelling fixes.
* Add missing docs.
* Queue tests.
* More tests.
* Add metrics + don't short circuit import.
* Basic test for ordering provider.
* Import fix.
* Remove dead link.
* One more dead link.
Co-authored-by: Lldenaurois <Ljdenaurois@gmail.com>
* statement-distribution: Only warn on relevant stuff.
* Silence warnings in availability-distribution.
* Demote more warnings.
* More consistency.
* info -> debug
* Attempt to add log stats to gossip-support.
* WIP: Keep track of connected validators.
* Clarify metric.
* WIP: Make gossip support report connectivity.
* WIP: Fixing tests.
* Fix network bridge + integrate in overseer.
* Consistent naming.
* Fix logic error
* cargo fmt
* Pretty logs.
* cargo fmt
* Use `Delay` to trigger periodic checks.
* fmt
* Fix warning for authority set size of 1.
* More correct ratio report if there are no resolved validators.
* Prettier rendering of empty set.
* Fix typo.
* Another typo.
* Don't check on every leaf update.
* Make compatible with older rustc.
* Fix tests.
* Demote warning.
* Better warning policy on dispute-distribution.
Network errors happen, only warn on a minimum number of failed attempts.
* Add metric to see how long it takes to get a dispute vote out.
* Let's be alert.
* Make SessionInfo include all authorities
of the current session.
* Add missing import.
* Only take subset for assignment keys.
* Fix typo.
* Make spellcheck happy.
* Really.
* Fix test.
* More clear documentation.
* Update comments in `SessionInfo`.
* WIP: Get rid of request multiplexer.
* WIP
* Receiver for handling of incoming requests.
* Get rid of useless `Fault` abstraction.
The things the type system let us do are not worth getting abstracted in
its own type. Instead error handling is going to be merely a pattern.
* Make most things compile again.
* Port availability distribution away from request multiplexer.
* Formatting.
* Port dispute distribution over.
* Fixup statement distribution.
* Handle request directly in collator protocol.
+ Only allow fatal errors at top level.
* Use direct request channel for availability recovery.
* Finally get rid of request multiplexer
Fixes#2842 and paves the way for more back pressure possibilities.
* Fix overseer and statement distribution tests.
* Fix collator protocol and network bridge tests.
* Fix tests in availability recovery.
* Fix availability distribution tests.
* Fix dispute distribution tests.
* Add missing dependency
* Typos.
* Review remarks.
* More remarks.
* Move on to next validator after timeout.
* Better naming.
* Wrong implementation of validator fetch timeouts.
* Validator side: Move on to next collator
if download takes too long.
* Drop multiple requests from same validator.
* Add test that next response is sent after timeout.
* Multiple requests by same validator should get dropped.
* Test that another collator is tried
after exclusive download time.
* Add dep.
* Cleanup.
* Merge fix.
* Review remarks.
* Fixes.
* Add log targets to trace logs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Dispute distribution initial design.
* WIP.
* Dispute distribution guide update.
* Make invalid statement include `InvalidStatementKind`.
* Clarify the scope of disputes.
* A few fixes + introduced back pressure oneshot.
* Fixes and spam protection WIP.
* More spam considerations.
* More fixes.
* Fixes + add note about not dispute participating nodes.
* State can be finished due to `Share` message.
Therefoe a task can still be running in that state. Removed panic and
changed state name to reflect possibility of `Share` message.
* bump spec versions in kusama, polkadot and westend again III
* properly bump for the upcoming release
Co-authored-by: Andronik Ordian <write@reusable.software>
* Remove signature verification in backing.
`SignedFullStatement` now signals that the signature has already been
checked.
* Remove unused check_payload function.
* Introduced unchecked signed variants.
* Fix inclusion to use unchecked variant.
* More unchecked variants.
* Use unchecked variants in protocols.
* Start fixing statement-distribution.
* Fixup statement distribution.
* Fix inclusion.
* Fix warning.
* Fix backing properly.
* Fix bitfield distribution.
* Make crypto store optional for `RuntimeInfo`.
* Factor out utility functions.
* get_group_rotation_info
* WIP: Collator cleanup + check signatures.
* Convenience signature checking functions.
* Check signature on collator-side.
* Fix warnings.
* Fix collator side tests.
* Get rid of warnings.
* Better Signed/UncheckedSigned implementation.
Also get rid of Encode/Decode for Signed! *party*
* Get rid of dead code.
* Move Signed in its own module.
* into_checked -> try_into_checked
* Fix merge.
* Factor out runtime module into utils.
* First fatal error design.
* Better error handling infra.
* Error handling cleanup.
* Send to peers of our group first.
* Finish backing group prioritization.
* Little cleanup.
* More cleanup.
* Forgot to checkin error.rs.
* Notes.
* Runtime -> RuntimeInfo
* qed in debug assert.
* PolkaErr -> Fault.
* Factor out runtime module into utils.
* Add maybe_authority information to `PeerConnected` event.
We already gather this information in authority discovery, so we might
as well share it with others.
This opens up an easy path to trigger validators differently from normal
nodes, e.g. for prioritization. This change has become more important
now, that we just connect to all validators and therefore just have a
long peer list without any information about those nodes.
* Test fix.
* More test coverage.
* Preserve peer order.
* Better test coverage.
* Even more test coverage.
* Add doc comment to `IndexMap`.
* Fix flaky test.
* Review remarks.
* Review remarks.
* Wip
* Increase proposer timeout.
* WIP.
* Better timeout values now that we are going to be connected to all nodes. (#2778)
* Better timeout values.
* Fix typo.
* Fix validator bandwidth.
* Fix compilation.
* Better and more consistent sizes.
Most importantly code size is now 5 Meg, which is the limit we currently
want to support in statement distribution.
* Introduce statement fetching request.
* WIP
* Statement cache retrieval logic.
* Review remarks by @rphmeier
* Fixes.
* Better requester logic.
* WIP: Handle requester messages.
* Missing dep.
* Fix request launching logic.
* Finish fetching logic.
* Sending logic.
* Redo code size calculations.
Now that max code size is compressed size.
* Update Cargo.lock (new dep)
* Get request receiver to statement distribution.
* Expose new functionality for responding to requests.
* Cleanup.
* Responder logic.
* Fixes + Cleanup.
* Cargo.lock
* Whitespace.
* Add lost copyright.
* Launch responder task.
* Typo.
* info -> warn
* Typo.
* Fix.
* Fix.
* Update comment.
* Doc fix.
* Better large statement heuristics.
* Fix tests.
* Fix network bridge tests.
* Add test for size estimate.
* Very simple tests that checks we get LargeStatement.
* Basic check, that fetching of large candidates is performed.
* More tests.
* Basic metrics for responder.
* More metrics.
* Use Encode::encoded_size().
* Some useful spans.
* Get rid of redundant metrics.
* Don't add peer on duplicate.
* Properly check hash
instead of relying on signatures alone.
* Preserve ordering + better flood protection.
* Get rid of redundant clone.
* Don't shutdown responder on failed query.
And add test for this.
* Smaller fixes.
* Quotes.
* Better queue size calculation.
* A bit saner response sizes.
* Fixes.
* availability-distribution: Retry on fail on next block.
Retry failed fetches on next block when still pending availability.
* Update node/network/availability-distribution/src/requester/fetch_task/mod.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Fix existing tests.
* Add test for trying all validators.
* Add test for testing retries.
Co-authored-by: Andronik Ordian <write@reusable.software>
* Remove stuff out of the runtime that does not belong there.
There might be more, but it is a start.
* White space fixes.
* Fix tests.
* Leave whitespace in ui tests alone.
* Add back zstd for no reason.
* Fix browser wasm (hopefully)
* Indentation fix.
* Prepare request-response for PoV fetching.
* Drop old PoV distribution.
* WIP: Fetch PoV directly from backing.
* Backing compiles.
* Runtime access and connection management for PoV distribution.
* Get rid of seemingly dead code.
* Implement PoV fetching.
Backing does not yet use it.
* Don't send `ConnectToValidators` for empty list.
* Even better - no need to check over and over again.
* PoV fetching implemented.
+ Typechecks
+ Should work
Missing:
- Guide
- Tests
- Do fallback fetching in case fetching from seconding validator fails.
* Check PoV hash upon reception.
* Implement retry of PoV fetching in backing.
* Avoid pointless validation spawning.
* Add jaeger span to pov requesting.
* Add back tracing.
* Review remarks.
* Whitespace.
* Whitespace again.
* Cleanup + fix tests.
* Log to log target in overseer.
* Fix more tests.
* Don't fail if group cannot be found.
* Simple test for PoV fetcher.
* Handle missing group membership better.
* Add test for retry functionality.
* Fix flaky test.
* Spaces again.
* Guide updates.
* Spaces.
* Those should really be trace.
- Very spammy
- And they in fact trace the execution
- Should not be enabled lightly - will slow network bridge down.
* Make report peers debug again.
* Bigger is better.
Made all request response sizes 10 times bigger.
* The smaller the better.
* Update comment.
* Ah, bigger is still better.
Max PoV size for rococo is around 50Meg, compression ratio is about 3.4.
With 30 Meg we should be fine, even with crypto kitties in the PoV.
Robert Klotzner