This PR allows sharing the pubkey data between validators by using a
thread-local cache for pubkey data, netting about a 400mb mem usage
reduction on holesky due to us keeping 3 permanent + several ephemeral
state copies in memory at all times and each state copy holding a full
validator.
The PR also introduces a hash cache for the key which gives ~14% speedup
for a full state `hash_tree_root` - the key makes up for a large part of
the `Validator` htr time.
Finally, the time it takes to copy a state goes down as well from ~80m
ms to ~60, for reasons similar to htr.
We use a `ptr` even if a `ref` could in theory have been used - there is
not much practical benefit to a `ref` (given it's mutable) while a `ptr`
is cheaper and easier to copy (when copying temporary states).
We could go further and cache a cooked pubkey but it turns out this is
quite intrusive - in all the relevant places, we're already using a
cooked key from the immutable validator data so there are no immediate
performance gains of doing so while managing the compressed -> cooked
key mapping would become more difficult - something for a future PR
perhaps.
Co-authored-by: Etan Kissling <etan@status.im>
* track latest duration instead of total in new timing metrics
Change `db_checkpoint_seconds` and `state_replay_seconds` metrics to
record the latest duration instead of the total. `nim-metrics` already
synthesizes a `_total` metric from these implicitly.
* still have to use inc, metrics only synthesizes the name not the sum
* prefix with `beacon_dag`
Validator monitoring gained 2 new metrics for tracking when blocks are
included or not on the head chain.
Similar to attestations, if the block is produced in epoch N, reporting
will use the state when switching to epoch N+2 to do the reporting (so
as to reasonably stabilise the block inclusion in the face of reorgs).
Database checkpointing can take seconds, e.g., while Geth is syncing.
Add a debug log + metric for it, and also info log if it takes longer
than 250ms, same as for the existing `State replayed` log. If the log
shows up for a user while the system is not overloaded, it may point
to slow disk speed or thermal issue.
* compute post-merge randao mix without loading state
* avoid copying state on shuffling computation and compute epochref
* speed up state copy for block production
With checkpoint sync, the checkpoint block is typically unavailable at
the start, and only backfilled later. To avoid treating it as having
zero hash, execution disabled in some contexts, wrap the result of
`loadExecutionBlockHash` in `Opt` and handle block hash being unknown.
---------
Co-authored-by: Jacek Sieka <jacek@status.im>
When syncing, we log a notice each time someone asks us for a block that
we haven't backfilled yet. This is quite verbose and not unexpected,
because the status message does not allow indicating backfill progress.
When using checkpoint sync, only checkpoint state is available, block is
not downloaded and backfilled later.
`dag.backfill` tracks latest filled `slot`, and latest `parent_root` for
which no block has been synced yet.
In checkpoint sync, this assumption is broken, because there, the start
`dag.backfill.slot` is set based on checkpoint state slot, and the block
is also not available.
However, sync manager in backward mode also requests `dag.backfill.slot`
and `block_clearance` then backfills the checkpoint block once it is
synced. But, there is no guarantee that a peer ever sends us that block.
They could send us all parent blocks and solely omit the checkpoint
block itself. In that situation, we would accept the parent blocks and
advance `dag.backfill`, and subsequently never request the checkpoint
block again, resulting in gap inside blocks DB that is never filled.
To mitigate that, the assumption is restored that `dag.backfill.slot`
is the latest filled `slot`, and `dag.backfill.parent_root` is the next
block that needs to be synced. By setting `slot` to `tail.slot + 1` and
`parent_root` to `tail.root`, we put a fake summary into `dag.backfill`
so that `block_clearance` only proceeds once checkpoint block exists.
After checkpoint sync, historical block IDs cannot yet be queried.
However, they are needed to compute dependent roots of `ShufflingRef`.
To allow lookup, enable `getBlockIdAtSlot` to answer from compatible
states in memory; as long as they descend from the finalized checkpoint
and the requested slot is sufficiently recent, `block_roots` contains
everything to recover `BlockSlotId` up to `SLOTS_PER_HISTORICAL_ROOT`.
This is similar to how `attester_dependent_root` etc. are computed.
This accelerates the first couple minutes of checkpoint sync on Mainnet,
especially the time until finality advances past the synced checkpoint.
Finish the rename started in #4809 to have a consistent naming.
`ExecutionPayloadHash` suggests hash over payload instead of block.
`BlockHash` is also the canonical name in engine API.
When using checkpoint sync, the initial block is missing in the DB.
Update the LC data collector initialization to account for that,
avoiding a spurious error message when it is incorrectly accessed:
```
ERR 2024-02-07 11:21:55.416+01:00 Block failed to load unexpectedly topics="chaindag_lc" bid=d30517a7:8257504 tail=8257504
```
Also fixes a regression from #5691 that resulted in similar messages
while importing the first few blocks after checkpoint sync.
Thanks to @arnetheduck for reporting this.
Full caches should not be used to mark blocks as unviable. The unviable
status is quite persistent and a block marked as such won't be processed
again once the cache empties. Problem originally introduced in #4808.
Changes here are more significant because of some good old tech debt in
block production which has grown quite hairy - the reduction in
exception handling at least provides some steps in the right direction.
`eth1_chain` no longer logs with `topics` since #5768, making it hard
to filter messages from this module. Re-add the `topics`, and also fix
outdated `topics` in `el_manager` (formerly `*_monitor`).
When BN clock is out of sync, VC sets BN status to `BrokenClock`. It is
only reset to `Offline` after restoring time sync. However, if VC fails
encounters an error while checking time, Nimbus extensions are assumed
to be unavailable and the BN is no longer checked for having a synced
clock. This means it is never reset back to `Offline` if errors start
occurring _after_ BN is already set to `BrokenClock`. This could be
because BN is changed from Nimbus to an alternative implementation,
or due to intermittent connection issues.
Ensure that BN status is reset back to `Offline` when Nimbus extensions
are disabled to ensure eventual connection recovery.
* Revert "use `RestPlainResponse` to improve builder API rerror reporting"
* Update rest_deneb_mev_calls.nim
copyright year linting
* Update rest_capella_mev_calls.nim
more copyright year linting
* Adopt asyncraises guarantees to most of the REST API handlers.
Bump presto.
* Fix copyright year.
---------
Co-authored-by: Etan Kissling <etan@status.im>
#5773 removed catching up on validator duties after lag. The `curSlot`
variable that was used originally to track catch-up progress no longer
has a use and is also no longer properly updated. Remove it.
If the initial state replays cover the finalized head, import matching
`LightClientBootstrap` into database.
This also addresses this error when light client requests bootstrap from
the genesis slot on networks that launch with Altair enabled.
```
{"lvl":"DBG","ts":"2023-10-04 11:17:49.665+00:00","msg":"LC bootstrap unavailable: Sync committee branch not cached","topics":"chaindag_lc","slot":0}
```
Avoid marking blocks invalid when corresponding `blobSidecarsByRange`
returns an incomplete / incorrect response while syncing. The block
itself may still be valid in that scenario.
There are two conditions leading to `duplicate contribution` log.
Align the logs with the ones used for attestation aggregates,
so that the two conditions can be separated when reading logs.
The `blob_gas_used` field was not properly populated when constructing
Deneb light client data. This is due to #5026 not applying the change to
the entire codebase when the new field got introduced, and due to #5350
not catching that oversight in other modules. Also reviewed codebase and
discovered that `shortLog` for Deneb execution payloads has same bug.
To simplify supporting "am I ready for the fork" requests, add a simple
marker to the default status bar that indicates readiness by displaying
the fork's corresponding mascot. This is the same one that is also
displayed during the actual fork transition, so does not introduce
new dependencies. It also only shows in default status bar, not in logs.
* remove extraneous length checks in KZG batch proofs
* re-add winservice import but only for Windows, to avoid UnusedImport warning
* also uses establishWindowsService
* Make some startup procedures async.
Add more handful makeBannerAndConfig().
* Dissect windows service code from `nimbus_beacon_node.nim`.
* Add report service startup errors using windows error codes.
Add plug able exitService().
Co-authored-by: Zahary Karadjov <zahary@status.im>
Co-authored-by: Jacek Sieka <jacek@status.im>
* Avoid global in p2p macro (fixes#4578)
* copy p2p macro to this repo and start de-crufting it
* make protocol registration dynamic, removing light client hacks et al
* split out light client protocol into its own file
* cleanups
* Option -> Opt
* remove more cruft
* further split beacon_sync
this allows the light client to respond to peer metadata messages
without exposing the block sync protocol
* better protocol init
* "constant" protocol index
* avoid casts
* copyright
* move some discovery code to discovery
* avoid extraneous data copy when sending chunks
* remove redundant forkdigest field
* document how to connect to a specific peer
Etan Kissling (2):
remove unused `skip0xPrefix`
keep the internal count helper
Will (1):
Bugfix/nully values (#61)
Yuriy Glukhov (5):
Contract constructor support
Fixed compilation error in exec function
Added string encoding
Fixed source->from field of EthCall
More flexibility to contract DSL, Async contract caller
jangko (5):
Reduce compiler warnings when using Nim v2
Migrate to json-serialization
Add tests of json rpc marshalled types
Resolve contract_dsl ambiguity
Event handler passing around JsonString instead of JsonNode
Share encoder between json-rpc and chronicles (#119)
Simplify generic constraint of rpc and chronicles encoders
Feature/execution api spec (#69)
v0.3.0
bump nim-json-rpc to a6475e49b26d3afc58aaa3d67621c94eafef8efb
coffeepots (1):
Use nim-json-serialization for RPCs (#172)
jangko (10):
Add copyright to source file
Remove StringOfJson
Fix optional parameter parsing fails in rpc macro with generics
Rename jrpc_sys module back to jsonmarshal
Reenable test hhtps
Add test for createRpcSigsFromNim and createSingleRpcSig
Let the OS choose the port for tests
Add onProcessMessage hook to client
Fix example in the README.md
Move errors module back to json_rpc folder
Upgrade rpc router internals (#178)
RPC server handle null return value correctly
v0.3.0
kdeme (1):
Add example test case that currently fails the Option parsing
Extend slot start message and default status bar with information about
current head fork and the next fork transition (corresponding to head).
This is useful to know whether a synced client is aware of a future fork
and can also be useful when syncing from old forks to follow progress
across the various forks.
```
peers: 8 ❯ finalized: 741c2ce2:230474 ❯ head: b330f58b:230477:20 ❯ fork: Capella (next: Deneb:231680) ❯ time: 230599:24 (7379192) ❯ sync: 00h24m (99.63%) 2.6492slots/s (QwQUwQPQDQ:7375263)/opt
```
```
INF 2024-01-12 12:18:00.001+01:00 Slot start topics="beacnde" slot=7379190 epoch=230599 fork="Capella (next: Deneb:231680)" sync="--h--m (99.62%) 0.0000slots/s (wwwwwwwwww:7375167)/opt" peers=0 head=741c2ce2:7375168 finalized=230472:723abe7e delay=1ms861us
```
In #5664, `nim-json-rpc` dependency got bumped which included a change
in behaviour when processing `null` data for heap allocated objects.
- https://github.com/status-im/nim-json-rpc/pull/176
Old behaviour was to raise an exception, while new behaviour is to set
the value to `nil` but treat it as a successful parse. Old exceptions
were similar to "Parameter [result] expected JObject but got JNull".
As part of the `nim-json-rpc` bump in #5664, `el_manager.nim` was not
updated to match the new behaviour, leading to crash whenever its logic
assumes that a successfully parsed web3 `BlockObject` (heap allocated)
may be assumed to be non-`nil`.
As a quick remedy, the `el_manager.nim` is updated to transform `nil`
responses for `BlockObject` into `ValueError`, allowing reuse of the
existing and tested exception based processing.
In #5664, `nim-json-rpc` dependency got bumped which included a change
in behaviour when processing `null` data for heap allocated objects.
- https://github.com/status-im/nim-json-rpc/pull/176
Old behaviour was to raise an exception, while new behaviour is to set
the value to `nil` but treat it as a successful parse. Old exceptions
were similar to "Parameter [result] expected JObject but got JNull".
As part of the `nim-json-rpc` bump in #5664, `el_manager.nim` was not
updated to match the new behaviour, leading to crash whenever its logic
assumes that a successfully parsed web3 `BlockObject` (heap allocated)
may be assumed to be non-`nil`.
As a quick remedy, the `el_manager.nim` is updated to transform `nil`
responses for `BlockObject` into `ValueError`, allowing reuse of the
existing and tested exception based processing.
This PR causes a few new warnings to appear - these are harmless but
will need addressing separately as they span several libraries.
* new asyncraises syntax
* asyncraises support in several modules
* `sink` usage reduces spurious copying
* `?` compatiblity for `async` + `results`
* remove `-d:chronosStrictException` (obsolete)
The `eth2-networks` repo often receives metadata updates with a delay.
Switch to `goerli` repo to obtain the latest config (Dencun scheduling)
when it is updated. This is in line with how Sepolia / Holesky work.
- https://github.com/eth-clients/goerli/pull/178
The `eth2-networks` repo often receives metadata updates with a delay.
Switch to `goerli` repo to obtain the latest config (Dencun scheduling)
when it is updated. This is in line with how Sepolia / Holesky work.
- https://github.com/eth-clients/goerli/pull/178
With Capella, `bls_to_execution_change` SSE should be emitted on the
event stream whenever a new `SignedBLSToExecutionChange` is received.
Add this missing functionality for compatibility with beacon-API specs.
- https://github.com/ethereum/beacon-APIs/pull/248
This requires all object types to be explicitly white-listed for
default serialization. The PR makes the minimal changes, although
a number of similar mechanisms in eth2_rest_serialization can now
be removed.
As followup from #5654, ensure that we still keep the comment around
referring to the correct `forkchoiceUpdated` to use being driven by
the fork schedule.
* use `PayloadAttributesV3` in `nimbus_light_client` for Deneb
From Deneb onward, `forkchoiceUpdated` requires `PayloadAttributesV3`.
In `nimbus_light_client` we still used `PayloadAttributesV2`.
Also clean up two other locations that were already correctly using
`PayloadAttributesV3`, to reduce code duplication.
* fix letter case
* avoid potentially subtle template/function symbol name interactions
* use warn instead of error in getExecutionPayload codepath to ensure lack of ambiguity
When the BN exits after writing new `head` to database, but before
completing the `updateFinalizedBlocks` call, the database is slightly
inconsistent due to the partial write. We currently fail to start up
after that. Fix that by catching up on partial `updateFinalizedBlocks`
tasks on start up, and add a test for this edge case.
Simplify best `LightClientUpdate` collection by tracking only canonical
data instead of tracking the best update across all branches within the
sync committee period.
- https://github.com/ethereum/consensus-specs/pull/3553
* reorder gossip validation checks
Doing the coverage check only after the corresponding committee index is
known allows optimization by early rejecting invalid data.
* use same helper for individual attestations as well
This PR brings down the time to send 100 attestations from ~1s to
~100ms, making it feasible to run 10k validators on a single node (which
regularly send 300 attestations / slot).
This is done by batching the slashing protection database write in a
single transaction thus avoiding a slow fsync for every signature -
effects will be more pronounced on slow drives.
The benefit applies both to beacon and client validators.
When creating new LC updates, information about the parent block's post
state must be available (cached), but information about current block's
post state is not yet required. Caching information about the current
block's post state can be delayed, simplifying the LC data collection
logic a bit and allowing more future flexibility with the cache design.
The proposed change ensures that VC validators are registered with the builder specified by the `--payload-builder-url` argument even if the beacon node has no attached validators. It also prevent such validators from being unintentionally registered with builders configured for specific attached validators by the keymanager api.
Per the current specs, the VC have no way to specify which builders the BN should use on a per-node basis, so for the time being we have to resort to using the BN fallback default builder URL for VC validators.
When new finality is reached without supermajority sync committee
support, trigger another event push on beacon-API and libp2p once
the finality gains supermajority support.
- https://github.com/ethereum/consensus-specs/pull/3549
Replace sections that need to be maintained with every `ConsensusFork`
related to LC data collection with a generic logic that keeps working
when unrelated parts of Ethereum change.
* Initial commit.
* Fix issues and tests.
* Fix test compilation issue.
* Update AllTests.
* Change the most poor score name from <lowest> to <bad>.
Split sync committee message score in range, so lexicographic scores will not intersect with normal one.
Lexicographic scores should be below to normal scores.
* Address review comments.
Fix aggregated attestation scoring to use MAX_VALIDATORS_PER_COMMITTEE.
Fix sync committee contributions to use SYNC_SUBCOMMITTEE_SIZE.
Add getUniqueVotes test vectors.
* Post-rebase fixes.
* Address review comments.
* Return back score calculation based on actual bits length.
* AllTests modification.
Gnosis uses `MIN_EPOCHS_FOR_BLOCK_REQUESTS` = 33024, but the computed
safe minimum (that Nimbus was using) is 2304. Relax the compatibility
check to allow `MIN_EPOCHS_FOR_BLOCK_REQUESTS` above the safe minimum
and honor `config.yaml` preferences for `MIN_EPOCHS_FOR_BLOCK_REQUESTS`.
* make `MIN_EPOCHS_FOR_BLOB_SIDECARS_REQUESTS` configurable
Gnosis uses custom `MIN_EPOCHS_FOR_BLOB_SIDECARS_REQUESTS` to account
for the faster slot timing, so that blobs still remain available for
roughly the same amount of real time.
Also extend REST config endpoint with full config form `v1.4.0-beta.4`,
and extend compatibility checks when loading configs to reduce warnings.
* Add slashing database pruning to VC.
Fix GetBlockHeaderResponse object declaration (spec has been changed).
* Switch to getFinalizedBlockHeader instead.
* Fix proper sign.
Add statements.
Show pruning log statement only when pruning happens.
* Optimize and remove debugging helpers.
tersec
Etan Kissling
Jacek Sieka
Zahary Karadjov
Kim De Mey
Eugene Kabanov
andri lim
Zahary Karadjov
stephanep
xiaolou86