* 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.
* add number to `ActivatedLeavesUpdate`
* update subsystem util and overseer
* use new ActivatedLeaf everywhere
* sort view
* sorted and limited view in network bridge
* use live block hash only if it's newer
* grumples
* Introduce collation fetching protocol
also move to mod.rs
* Allow `PeerId`s in requests to network bridge.
* Fix availability distribution tests.
* Move CompressedPoV to primitives.
* Request based collator protocol: validator side
- Missing: tests
- Collator side
- don't connect, if not connected
* Fixes.
* Basic request based collator side.
* Minor fix on collator side.
* Don't connect in requests in collation protocol.
Also some cleanup.
* Fix PoV distribution
* Bump substrate
* Add back metrics + whitespace fixes.
* Add back missing spans.
* More cleanup.
* Guide update.
* Fix tests
* Handle results in tests.
* Fix weird compilation issue.
* Add missing )
* Get rid of dead code.
* Get rid of redundant import.
* Fix runtime build.
* Cleanup.
* Fix wasm build.
* Format fixes.
Thanks @andronik !
* Logging functionality for spans.
* Jaeger spans for availability distribution.
* Fix instrumentation to use log target properly.
* Add some tracing instrumentation macros.
* Use int_tags instead of logs.
* Add span per iteration.
* Remove span::log functionality.
* Fix instrumentation log target for real.
* Add jaeger span to responding side as well.
* Revert "Fix instrumentation log target for real."
This reverts commit e1c2a2e6ff6f257e702f07d8a77c2668af92b0ef.
* Revert "Fix instrumentation to use log target properly."
This reverts commit 7caa0bd1acc6fe9727bb3a91851560d756c40ab8.
* target -> subsystem in instrumentatio macro
target is not correct either, and the correct way of using a top level
target = LOG_TARGET does not work, as the macro expects a string literal
and gets confused by the constant `LOG_TARGET`.
* Use kebab-case for spa names.
Co-authored-by: Andronik Ordian <write@reusable.software>
Co-authored-by: Andronik Ordian <write@reusable.software>
* WIP: Whole subsystem test.
* New tests compile.
* Avoid needless runtime queries for no validator nodes.
* Make tx and rx publicly accessible in virtual overseer.
This simplifies mocking in some cases, as tx can be cloned, but rx can
not.
* Whole subsystem test working.
* Update node/network/availability-distribution/src/session_cache.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Update node/network/availability-distribution/src/session_cache.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Document better what `None` return value means.
* Get rid of BitVec dependency.
* Update Cargo.lock
* Hopefully fixed implementers guide build.
Co-authored-by: Andronik Ordian <write@reusable.software>
* WIP
* availability distribution, still very wip.
Work on the requesting side of things.
* Some docs on what I intend to do.
* Checkpoint of session cache implementation
as I will likely replace it with something smarter.
* More work, mostly on cache
and getting things to type check.
* Only derive MallocSizeOf and Debug for std.
* availability-distribution: Cache feature complete.
* Sketch out logic in `FetchTask` for actual fetching.
- Compile fixes.
- Cleanup.
* Format cleanup.
* More format fixes.
* Almost feature complete `fetch_task`.
Missing:
- Check for cancel
- Actual querying of peer ids.
* Finish FetchTask so far.
* Directly use AuthorityDiscoveryId in protocol and cache.
* Resolve `AuthorityDiscoveryId` on sending requests.
* Rework fetch_task
- also make it impossible to check the wrong chunk index.
- Export needed function in validator_discovery.
* From<u32> implementation for `ValidatorIndex`.
* Fixes and more integration work.
* Make session cache proper lru cache.
* Use proper lru cache.
* Requester finished.
* ProtocolState -> Requester
Also make sure to not fetch our own chunk.
* Cleanup + fixes.
* Remove unused functions
- FetchTask::is_finished
- SessionCache::fetch_session_info
* availability-distribution responding side.
* Cleanup + Fixes.
* More fixes.
* More fixes.
adder-collator is running!
* Some docs.
* Docs.
* Fix reporting of bad guys.
* Fix tests
* Make all tests compile.
* Fix test.
* Cleanup + get rid of some warnings.
* state -> requester
* Mostly doc fixes.
* Fix test suite.
* Get rid of now redundant message types.
* WIP
* Rob's review remarks.
* Fix test suite.
* core.relay_parent -> leaf for session request.
* Style fix.
* Decrease request timeout.
* Cleanup obsolete errors.
* Metrics + don't fail on non fatal errors.
* requester.rs -> requester/mod.rs
* Panic on invalid BadValidator report.
* Fix indentation.
* Use typed default timeout constant.
* Make channel size 0, as each sender gets one slot anyways.
* Fix incorrect metrics initialization.
* Fix build after merge.
* More fixes.
* Hopefully valid metrics names.
* Better metrics names.
* Some tests that already work.
* Slightly better docs.
* Some more tests.
* Fix network bridge test.
* feat/view: assure heads in a view are sorted
Allows O(n) comparisons, adds an alternate equiv relation
which takes O(n^2) for integrity verification.
Ref #2133
* revert: remove custom PartialEq impl, there are no duplicates
* fix: do not sort the live_heads, that alters the local view
* refactor/view: heads should not be public
* chore/spellcheck: add unfinalized
* fix/view: add missing len() and is_empty() fns
* quirk
* vec is not view
* Update node/network/approval-distribution/src/tests.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Update node/network/bridge/src/lib.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Update node/network/protocol/src/lib.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* fixup comment
* fix botched test
Co-authored-by: Andronik Ordian <write@reusable.software>
* refactor/reputation: unify the values used
* chore/rep: rename Annoy* to Cost*, make duplicate message Cost*Repeated
* fix/reputation: lost and found, convert at the boundary to substrate
* refactor/rep: move conversion to base reputation one level down, left conversions
* fix/rep: order of magnitude adjustments
Thanks pierre!
* remove spaces
* chore/rep: give rationale for order of magnitude
* refactor/rep: move UnifiedReputationChange to separate file
* fix/rep: order of magnitudes correction
* Move NetworkBridgeEvent to subsystem::messages.
It is not protocol related at all, it is in fact only part of the
subsystem communication as it gets wrapped into messages of each
subsystem.
* Request/response infrastructure is taking shape.
WIP: Does not compile.
* Multiplexer variant not supported by Rusts type system.
* request_response::request type checks.
* Cleanup.
* Minor fixes for request_response.
* Implement request sending + move multiplexer.
Request multiplexer is moved to bridge, as there the implementation is
more straight forward as we can specialize on `AllMessages` for the
multiplexing target.
Sending of requests is mostly complete, apart from a few `From`
instances. Receiving is also almost done, initializtion needs to be
fixed and the multiplexer needs to be invoked.
* Remove obsolete multiplexer.
* Initialize bridge with multiplexer.
* Finish generic request sending/receiving.
Subsystems are now able to receive and send requests and responses via
the overseer.
* Doc update.
* Fixes.
* Link issue for not yet implemented code.
* Fixes suggested by @ordian - thanks!
- start encoding at 0
- don't crash on zero protocols
- don't panic on not yet implemented request handling
* Update node/network/protocol/src/request_response/v1.rs
Use index 0 instead of 1.
Co-authored-by: Andronik Ordian <write@reusable.software>
* Update node/network/protocol/src/request_response.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Fix existing tests.
* Better avoidance of division by zoro errors.
* Doc fixes.
* send_request -> start_request.
* Fix missing renamings.
* Update substrate.
* Pass TryConnect instead of true.
* Actually import `IfDisconnected`.
* Fix wrong import.
* Update node/network/bridge/src/lib.rs
typo
Co-authored-by: Pierre Krieger <pierre.krieger1708@gmail.com>
* Update node/network/bridge/src/multiplexer.rs
Remove redundant import.
Co-authored-by: Pierre Krieger <pierre.krieger1708@gmail.com>
* Stop doing tracing from within `From` instance.
Thanks for the catch @tomaka!
* Get rid of redundant import.
* Formatting cleanup.
* Fix tests.
* Add link to issue.
* Clarify comments some more.
* Fix tests.
* Formatting fix.
* tabs
* Fix link
Co-authored-by: Bernhard Schuster <bernhard@ahoi.io>
* Use map_err.
Co-authored-by: Bernhard Schuster <bernhard@ahoi.io>
* Improvements inspired by suggestions by @drahnr.
- Channel size is now determined by function.
- Explicitely scope NetworkService::start_request.
Co-authored-by: Andronik Ordian <write@reusable.software>
Co-authored-by: Pierre Krieger <pierre.krieger1708@gmail.com>
Co-authored-by: Bernhard Schuster <bernhard@ahoi.io>
* PVD: `block_number`->`relay_parent_number`
* ValidationParams: `relay_chain_height`->`relay_parent_number`
* Expose DMQ MQC hash as a well-known-key
This way the relay storage merkle proofs will be able to obtain the DMQ
MQC hash and we will be able to remove the it from the
PersistedValidationData struct.
* PersistedValidationData: Remove HRMP MQC heads
* PersistedValidationData: Remove `dmq_mqc_head`
* Expose the HRMP ingress channel index as a well-known-key
This way a parachain (PVF and collator) can find all the parachains that
have an outbound channel to the given one. That allows in turn to find
all the inbound channels for the given para.
Having access to that allows the parachain to get the same information
as the hrmp_mqc_heads now provide.
* Rename `relay_storage_root` to `relay_parent_storage_root`
* Do not send empty view updates to peers
It happened that we send empty view updates to our peers, because we
only updated our finalized block. This could lead to situations where we
overwhelmed sub systems with too many messages. On Rococo this lead to
constant restarts of our nodes, because some node apparently was
finalizing a lot of blocks.
To prevent this, the pr is doing the following:
1. If a peer sends us an empty view, we report this peer and decrease it
reputation.
2. We ensure that we only send a view update when the `heads` changed
and not only the `finalized_number`.
3. We do not send empty `ActiveLeavesUpdates` from the overseer, as this
makes no sense to send these empty updates. If some subsystem is relying
on the finalized block, it needs to listen for the overseer signal.
* Update node/network/bridge/src/lib.rs
Co-authored-by: Peter Goodspeed-Niklaus <coriolinus@users.noreply.github.com>
* Don't work if they're are no added heads
* Fix test
* Ahhh
* More fixes
Co-authored-by: Peter Goodspeed-Niklaus <coriolinus@users.noreply.github.com>
* Store all chunks and in a single transaction
* Adds chunks LRU to store
* Add pruning records metrics
* Use honest cache instead of LRU
* Remove unnecessary optional cache
* Fix review nits that are fixable
* Add one Jaeger span per relay parent
This adds one Jaeger span per relay parent, instead of always creating
new spans per relay parent. This should improve the UI view, because
subsystems are now grouped below one common span.
* Fix doc tests
* Replace `PerLeaveSpan` to `PerLeafSpan`
* More renaming
* Moare
* Update node/subsystem/src/lib.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* Skip the spans
* Increase `spec_version`
Co-authored-by: Andronik Ordian <write@reusable.software>
* Cont.: Implement the state root obtaining during inclusion
During inclusion now we obtain the storage root by passing it through
the inclusion_inherent.
* Fix tests
* Bump rococo spec version
* Reorder the parent header into the end
of the inclusion inherent.
When the parent header is in the beginning, it shifts the other two
fields, so that a previous version won't be able to decode that. If
we put the parent header in the end, the other two fields will stay
at their positions, thus make it possible to decode with the previous
version.
That allows us to perform upgrade of rococo runtime without needing of
simultanuous upgrade of nodes and runtime, or restart of the network.
* Squash a stray tab
* guide: add candidate information to OccupiedCore
* add descriptor and hash to occupied core type
* guide: add candidate hash to inclusion
* runtime: return candidate info in core state
* bitfield signing: stop querying runtime as much
* minimize going to runtime in availability distribution
* fix availability distribution tests
* guide: remove para ID from Occupied core
* get all crates compiling
* Fix bug and further optimizations in availability distribution
- There was a bug that resulted in only getting one candidate per block
as the candidates were put into the hashmap with the relay block hash as
key. The solution for this is to use the candidate hash and the relay
block hash as key.
- We stored received/sent messages with the candidate hash and chunk
index as key. The candidate hash wasn't required in this case, as the
messages are already stored per candidate.
* Update node/core/bitfield-signing/src/lib.rs
Co-authored-by: Robert Habermeier <rphmeier@gmail.com>
* Remove the reverse map
* major refactor of receipts & query_live
* finish refactoring
remove ancestory mapping,
improve relay-parent cleanup & receipts-cache cleanup,
add descriptor to `PerCandidate`
* rename and rewrite query_pending_availability
* add a bunch of consistency tests
* Add some last changes
* xy
* fz
* Make it compile again
* Fix one test
* Fix logging
* Remove some buggy code
* Make tests work again
* Move stuff around
* Remove dbg
* Remove state from test_harness
* More refactor and new test
* New test and fixes
* Move metric
* Remove "duplicated code"
* Fix tests
* New test
* Change break to continue
* Update node/core/av-store/src/lib.rs
* Update node/core/av-store/src/lib.rs
* Update node/core/bitfield-signing/src/lib.rs
Co-authored-by: Fedor Sakharov <fedor.sakharov@gmail.com>
* update guide to match live_candidates changes
* add comment
* fix bitfield signing
Co-authored-by: Robert Habermeier <rphmeier@gmail.com>
Co-authored-by: Bernhard Schuster <bernhard@ahoi.io>
Co-authored-by: Fedor Sakharov <fedor.sakharov@gmail.com>
* refactor View to include finalized_number
* guide: update the NetworkBridge on BlockFinalized
* av-store: fix the tests
* actually fix tests
* grumbles
* ignore macro doctest
* use Hash::repeat_bytes more consistently
* broadcast empty leaves updates as well
* fix issuing view updates on empty leaves updates
* use snake_case for log targets
* remove unused continue
* validator_discovery: when disconnecting, use all addresses
* validator_discovery: simplify request revokation
* fix a typo
* reexport prometheus-super for ease of use of other subsystems
* add some prometheus timers for collation generation subsystem
* add timing metrics to av-store
* add metrics to candidate backing
* add timing metric to bitfield signing
* add timing metrics to candidate selection
* add timing metrics to candidate-validation
* add timing metrics to chain-api
* add timing metrics to provisioner
* add timing metrics to runtime-api
* add timing metrics to availability-distribution
* add timing metrics to bitfield-distribution
* add timing metrics to collator protocol: collator side
* add timing metrics to collator protocol: validator side
* fix candidate validation test failures
* add timing metrics to pov distribution
* add timing metrics to statement-distribution
* use substrate_prometheus_endpoint prometheus reexport instead of prometheus_super
* don't include JOB_DELAY in bitfield-signing metrics
* give adder-collator ability to easily export its genesis-state and validation code
* wip: adder-collator pushbutton script
* don't attempt to register the adder-collator automatically
Instead, get these values with
```sh
target/release/adder-collator export-genesis-state
target/release/adder-collator export-genesis-wasm
```
And then register the parachain on https://polkadot.js.org/apps/?rpc=ws%3A%2F%2F127.0.0.1%3A9944#/explorer
To collect prometheus data, after running the script, create `prometheus.yml` per the instructions
at https://www.notion.so/paritytechnologies/Setting-up-Prometheus-locally-835cb3a9df7541a781c381006252b5ff
and then run:
```sh
docker run -v `pwd`/prometheus.yml:/etc/prometheus/prometheus.yml:z --network host prom/prometheus
```
Demonstrates that data makes it across to prometheus, though it is likely to be useful in the future
to tweak the buckets.
* Update parachain/test-parachains/adder/collator/src/cli.rs
Co-authored-by: Andronik Ordian <write@reusable.software>
* use the grandpa-pause parameter
* skip metrics in tracing instrumentation
* remove unnecessary grandpa_pause cli param
Co-authored-by: Andronik Ordian <write@reusable.software>
* drop in tracing to replace log
* add structured logging to trace messages
* add structured logging to debug messages
* add structured logging to info messages
* add structured logging to warn messages
* add structured logging to error messages
* normalize spacing and Display vs Debug
* add instrumentation to the various 'fn run'
* use explicit tracing module throughout
* fix availability distribution test
* don't double-print errors
* remove further redundancy from logs
* fix test errors
* fix more test errors
* remove unused kv_log_macro
* fix unused variable
* add tracing spans to collation generation
* add tracing spans to av-store
* add tracing spans to backing
* add tracing spans to bitfield-signing
* add tracing spans to candidate-selection
* add tracing spans to candidate-validation
* add tracing spans to chain-api
* add tracing spans to provisioner
* add tracing spans to runtime-api
* add tracing spans to availability-distribution
* add tracing spans to bitfield-distribution
* add tracing spans to network-bridge
* add tracing spans to collator-protocol
* add tracing spans to pov-distribution
* add tracing spans to statement-distribution
* add tracing spans to overseer
* cleanup
We need to distribute the PoV after we have seconded it. Other nodes
that will receive our `Secondded` statement and want to validate the
candidate another time will request this PoV from us.
* Make `CandidateHash` a real type
This pr adds a new type `CandidateHash` that is used instead of the
opaque `Hash` type. This helps to ensure on the type system level that
we are passing the correct types.
This pr also fixes wrong usage of `relay_parent` as `candidate_hash`
when communicating with the av storage.
* Update core-primitives/src/lib.rs
Co-authored-by: Peter Goodspeed-Niklaus <coriolinus@users.noreply.github.com>
* Wrap the lines
Co-authored-by: Peter Goodspeed-Niklaus <coriolinus@users.noreply.github.com>
* DMP: data structures and plumbing
* DMP: Implement DMP logic in the router module
DMP: Integrate DMP parts into the inclusion module
* DMP: Introduce the max size limit for the size of a downward message
* DMP: Runtime API for accessing inbound messages
* OCD
Small clean ups
* DMP: fix the naming of the error
* DMP: add caution about a non-existent recipient
* start working on building the real overseer
Unfortunately, this fails to compile right now due to an upstream
failure to compile which is probably brought on by a recent upgrade
to rustc v1.47.
* fill in AllSubsystems internal constructors
* replace fn make_metrics with Metrics::attempt_to_register
* update to account for #1740
* remove Metrics::register, rename Metrics::attempt_to_register
* add 'static bounds to real_overseer type params
* pass authority_discovery and network_service to real_overseer
It's not straightforwardly obvious that this is the best way to handle
the case when there is no authority discovery service, but it seems
to be the best option available at the moment.
* select a proper database configuration for the availability store db
* use subdirectory for av-store database path
* apply Basti's patch which avoids needing to parameterize everything on Block
* simplify path extraction
* get all tests to compile
* Fix Prometheus double-registry error
for debugging purposes, added this to node/subsystem-util/src/lib.rs:472-476:
```rust
Some(registry) => Self::try_register(registry).map_err(|err| {
eprintln!("PrometheusError calling {}::register: {:?}", std::any::type_name::<Self>(), err);
err
}),
```
That pointed out where the registration was failing, which led to
this fix. The test still doesn't pass, but it now fails in a new
and different way!
* authorities must have authority discovery, but not necessarily overseer handlers
* fix broken SpawnedSubsystem impls
detailed logging determined that using the `Box::new` style of
future generation, the `self.run` method was never being called,
leading to dropped receivers / closed senders for those subsystems,
causing the overseer to shut down immediately.
This is not the final fix needed to get things working properly,
but it's a good start.
* use prometheus properly
Prometheus lets us register simple counters, which aren't very
interesting. It also allows us to register CounterVecs, which are.
With a CounterVec, you can provide a set of labels, which can
later be used to filter the counts.
We were using them wrong, though. This pattern was repeated in a
variety of places in the code:
```rust
// panics with an cardinality mismatch
let my_counter = register(CounterVec::new(opts, &["succeeded", "failed"])?, registry)?;
my_counter.with_label_values(&["succeeded"]).inc()
```
The problem is that the labels provided in the constructor are not
the set of legal values which can be annotated, but a set of individual
label names which can have individual, arbitrary values.
This commit fixes that.
* get av-store subsystem to actually run properly and not die on first signal
* typo fix: incomming -> incoming
* don't disable authority discovery in test nodes
* Fix rococo-v1 missing session keys
* Update node/core/av-store/Cargo.toml
* try dummying out av-store on non-full-nodes
* overseer and subsystems are required only for full nodes
* Reduce the amount of warnings on browser target
* Fix two more warnings
* InclusionInherent should actually have an Inherent module on rococo
* Ancestry: don't return genesis' parent hash
* Update Cargo.lock
* fix broken test
* update test script: specify chainspec as script argument
* Apply suggestions from code review
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
* Update node/service/src/lib.rs
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
* node/service/src/lib: Return error via ? operator
* post-merge blues
* add is_collator flag
* prevent occasional av-store test panic
* simplify fix; expand application
* run authority_discovery in Role::Discover when collating
* distinguish between proposer closed channel errors
* add IsCollator enum, remove is_collator CLI flag
* improve formatting
* remove nop loop
* Fix some stuff
Co-authored-by: Andronik Ordian <write@reusable.software>
Co-authored-by: Bastian Köcher <git@kchr.de>
Co-authored-by: Fedor Sakharov <fedor.sakharov@gmail.com>
Co-authored-by: Robert Habermeier <robert@Roberts-MBP.lan1>
Co-authored-by: Bastian Köcher <bkchr@users.noreply.github.com>
Co-authored-by: Max Inden <mail@max-inden.de>
* Initial commit
* Move tests to separate module
* Move timestamps to the newtype
* Change idx name
* Use Duration for consts and update chunk records
* Ordering::Equal
* Update node/core/av-store/src/lib.rs
Co-authored-by: Peter Goodspeed-Niklaus <coriolinus@users.noreply.github.com>
* put_ methods do the array sorting
* Fix get_block_number
* Change StoreChunk message type and relay parent method
* Add chunk tests
* Fix block number computation for StoreChunk
* Duration instead of u64 everywhere
* Add a clarifying comment
Co-authored-by: Peter Goodspeed-Niklaus <coriolinus@users.noreply.github.com>
Robert Klotzner