Validator monitoring improves logging by giving more specific monitoring
information, and can now be seen as complete.
Previously, logging has focused on "Attestation sent" messages which
carry little informational value when things go wrong, and are overly
aggressive when everything works as expected (sending attestations is
the norm).
* lower "Attestation sent" log to `INFO`
* mark 1.7.0 as the start of the validator monitor feature - previous
versions had significant bugs in totals mode
* harden validator API against pre-finalized slot requests
* check `syncHorizon` when responding to validator api requests too far
from `head`
* limit state-id based requests to one epoch ahead of `head`
* put historic data bounds on block/attestation/etc validator production API, preventing them from being used with already-finalized slots
* add validator block smoke tests
* make rest test create a new genesis with the tests running roughly in
the first epoch to allow testing a few more boundary conditions
Recently, block processing times have been going up as the network grows
making early attestation riskier. Since blocks are big and attestations
are small (though numerous and therefore bandwidth-intense), it seems
better to wait a little bit longer after receiving a block, before we
publish the attestation.
Up til now, the block dag has been using `BlockRef`, a structure adapted
for a full DAG, to represent all of chain history. This is a correct and
simple design, but does not exploit the linearity of the chain once
parts of it finalize.
By pruning the in-memory `BlockRef` structure at finalization, we save,
at the time of writing, a cool ~250mb (or 25%:ish) chunk of memory
landing us at a steady state of ~750mb normal memory usage for a
validating node.
Above all though, we prevent memory usage from growing proportionally
with the length of the chain, something that would not be sustainable
over time - instead, the steady state memory usage is roughly
determined by the validator set size which grows much more slowly. With
these changes, the core should remain sustainable memory-wise post-merge
all the way to withdrawals (when the validator set is expected to grow).
In-memory indices are still used for the "hot" unfinalized portion of
the chain - this ensure that consensus performance remains unchanged.
What changes is that for historical access, we use a db-based linear
slot index which is cache-and-disk-friendly, keeping the cost for
accessing historical data at a similar level as before, achieving the
savings at no percievable cost to functionality or performance.
A nice collateral benefit is the almost-instant startup since we no
longer load any large indicies at dag init.
The cost of this functionality instead can be found in the complexity of
having to deal with two ways of traversing the chain - by `BlockRef` and
by slot.
* use `BlockId` instead of `BlockRef` where finalized / historical data
may be required
* simplify clearance pre-advancement
* remove dag.finalizedBlocks (~50:ish mb)
* remove `getBlockAtSlot` - use `getBlockIdAtSlot` instead
* `parent` and `atSlot` for `BlockId` now require a `ChainDAGRef`
instance, unlike `BlockRef` traversal
* prune `BlockRef` parents on finality (~200:ish mb)
* speed up ChainDAG init by not loading finalized history index
* mess up light client server error handling - this need revisiting :)
One more step on the journey to reduce `BlockRef` usage across the
codebase - this one gets rid of `StateData` whose job was to keep track
of which block was last assigned to a state - these duties have now been
taken over by `latest_block_root`, a fairly recent addition that
computes this block root from state data (at a small cost that should be
insignificant)
99% mechanical change.
* fewer deps on `BlockRef` traversal in anticipation of pruning
* allows identifying EpochRef:s by their shuffling as a first step of
* tighten error handling around missing blocks
using the zero hash for signalling "missing block" is fragile and easy
to miss - with checkpoint sync now, and pruning in the future, missing
blocks become "normal".
* Initial commit.
* Fix current test suite.
* Fix keymanager api test.
* Fix wss_sim.
* Add more keystore_management tests.
* Recover deleted isEmptyDir().
* Add `HttpHostUri` distinct type.
Move keymanager calls away from rest_beacon_calls to rest_keymanager_calls.
Add REST serialization of RemoteKeystore and Keystore object.
Add tests for Remote Keystore management API.
Add tests for Keystore management API (Add keystore).
Fix serialzation issues.
* Fix test to use HttpHostUri instead of Uri.
* Add links to specification in comments.
* Remove debugging echoes.
* deactivate Doppelganger Protection during genesis
* also don't actually flag supposed-doppelgangers (because they're before broadcastStartEpoch) on GENESIS_SLOT start
* update action tracker on dependent-root-changing reorg (instead of
epoch change)
* don't try to log duties while syncing - we're not tracking actions yet
* fix slot used for doppelganger loss detection
These use a separate flow, and were previously only registered from the
network
* don't log successes in totals mode (TMI)
* remove `attestation-sent` event which is unused
The current counters set gauges etc to the value of the _last_ validator
to be processed - as the name of the feature implies, we should be using
sums instead.
* fix missing beacon state metrics on startup, pre-first-head-selection
* fix epoch metrics not being updated on cross-epoch reorg
* limit by-root requests to non-finalized blocks
Presently, we keep a mapping from block root to `BlockRef` in memory -
this has simplified reasoning about the dag, but is not sustainable with
the chain growing.
We can distinguish between two cases where by-root access is useful:
* unfinalized blocks - this is where the beacon chain is operating
generally, by validating incoming data as interesting for future fork
choice decisions - bounded by the length of the unfinalized period
* finalized blocks - historical access in the REST API etc - no bounds,
really
In this PR, we limit the by-root block index to the first use case:
finalized chain data can more efficiently be addressed by slot number.
Future work includes:
* limiting the `BlockRef` horizon in general - each instance is 40
bytes+overhead which adds up - this needs further refactoring to deal
with the tail vs state problem
* persisting the finalized slot-to-hash index - this one also keeps
growing unbounded (albeit slowly)
Anyway, this PR easily shaves ~128mb of memory usage at the time of
writing.
* No longer honor `BeaconBlocksByRoot` requests outside of the
non-finalized period - previously, Nimbus would generously return any
block through this libp2p request - per the spec, finalized blocks
should be fetched via `BeaconBlocksByRange` instead.
* return `Opt[BlockRef]` instead of `nil` when blocks can't be found -
this becomes a lot more common now and thus deserves more attention
* `dag.blocks` -> `dag.forkBlocks` - this index only carries unfinalized
blocks from now - `finalizedBlocks` covers the other `BlockRef`
instances
* in backfill, verify that the last backfilled block leads back to
genesis, or panic
* add backfill timings to log
* fix missing check that `BlockRef` block can be fetched with
`getForkedBlock` reliably
* shortcut doppelganger check when feature is not enabled
* in REST/JSON-RPC, fetch blocks without involving `BlockRef`
* fix dag.blocks ref
* initial support for minification and new interchange tests. Removal of v1 and v1 migration.
* Synthetic attestations: SQLite3 requires one statement/query per prepared statement
* Fix DB import interrupted if no attestation was found
* Skip test relying on undocumented test behavior (https://github.com/eth-clients/slashing-protection-interchange-tests/pull/12#issuecomment-1011158701)
* Skip test relying on unclear minification behavior:
creating an invalid minified attestation with source > target or setting target = max(source, target)
* remove DB v1 and update submodule
* Apply suggestions from code review
Co-authored-by: Jacek Sieka <jacek@status.im>
Co-authored-by: Jacek Sieka <jacek@status.im>
Time in the beacon chain is expressed relative to the genesis time -
this PR creates a `beacon_time` module that collects helpers and
utilities for dealing the time units - the new module does not deal with
actual wall time (that's remains in `beacon_clock`).
Collecting the time related stuff in one place makes it easier to find,
avoids some circular imports and allows more easily identifying the code
actually needs wall time to operate.
* move genesis-time-related functionality into `spec/beacon_time`
* avoid using `chronos.Duration` for time differences - it does not
support negative values (such as when something happens earlier than it
should)
* saturate conversions between `FAR_FUTURE_XXX`, so as to avoid
overflows
* fix delay reporting in validator client so it uses the expected
deadline of the slot, not "closest wall slot"
* simplify looping over the slots of an epoch
* `compute_start_slot_at_epoch` -> `start_slot`
* `compute_epoch_at_slot` -> `epoch`
A follow-up PR will (likely) introduce saturating arithmetic for the
time units - this is merely code moves, renames and fixing of small
bugs.
* Harden CommitteeIndex, SubnetId, SyncSubcommitteeIndex
Harden the use of `CommitteeIndex` et al to prevent future issues by
using a distinct type, then validating before use in several cases -
datatypes in spec are kept simple though so that invalid data still can
be read.
* fix invalid epoch used in REST
`/eth/v1/beacon/states/{state_id}/committees` committee length (could
return invalid data)
* normalize some variable names
* normalize committee index loops
* fix `RestAttesterDuty` to use `uint64` for `validator_committee_index`
* validate `CommitteeIndex` on ingress in REST API
* update rest rules with stricter parsing
* better REST serializers
* save lots of memory by not using `zip` ...at least a few bytes!
With checkpoint sync in particular, and state pruning in the future,
loading states or state-dependent data may fail. This PR adjusts the
code to allow this to be handled gracefully.
In particular, the new availability assumption is that states are always
available for the finalized checkpoint and newer, but may fail for
anything older.
The `tail` remains the point where state loading de-facto fails, meaning
that between the tail and the finalized checkpoint, we can still get
historical data (but code should be prepared to handle this as an
error).
However, to harden the code against long replays, several operations
which are assumed to work only with non-final data (such as gossip
verification and validator duties) now limit their search horizon to
post-finalized data.
* harden several state-dependent operations by logging an error instead
of introducing a panic when state loading fails
* `withState` -> `withUpdatedState` to differentiate from the other
`withState`
* `updateStateData` can now fail if no state is found in database - it
is also hardened against excessively long replays
* `getEpochRef` can now fail when replay fails
* reject blocks with invalid target root - they would be ignored
previously
* fix recursion bug in `isProposed`
* log doppelganger detection when it activates and when it causes missed
duties
* less prominent eth1 sync progress
* log in-progress sync at notice only when actually missing duties
* better detail in replay log
* don't log finalization checkpoints - this is quite verbose when
syncing and already included in "Slot start"
Validator monitoring based on and mostly compatible with the
implementation in Lighthouse - tracks additional logs and metrics for
specified validators so as to stay on top on performance.
The implementation works more or less the following way:
* Validator pubkeys are singled out for monitoring - these can be
running on the node or not
* For every action that the validator takes, we record steps in the
process such as messages being seen on the network or published in the
API
* When the dust settles at the end of an epoch, we report the
information from one epoch before that, which coincides with the
balances being updated - this is a tradeoff between being correct
(waiting for finalization) and providing relevant information in a
timely manner)
* batch-verify sync messages for a small perf boost
Generally reuses the same structure as attestation and aggregate
verification
* normalize `signatures` and `signature_batch` to use the same pattern
of verification
* normalize parameter names, order etc for signature stuff in general
* avoid calling `blsSign` directly - instead, go through `signatures`
consistently
This PR fixes two issues with block publishing:
* Gossip-valid blocks are published before integrating them into the
chain, giving broadcasting a head start, both for rest block and
* Outright invalid blocks from the API that could lead to the descoring
of the node are no longer broadcast
Bonus:
* remove undocumented and duplicated `post_v1_validator_block` JSON-RPC
call
Validator clients such as Vouch can be configured to work with multiple
beacon nodes simultaneously. In this configuration, the validator client
will try to broadcast the gossip messages through each of the connected
beacon nodes which may lead to a situation where some of the nodes see a
message arriving from the network before it arrives through the REST API.
This should not be considered an error and the beacon node should still
broadcast the message as the intented purpose of the Vouch strategy is
to ensure that the message will reach as many peers as possible.
Renames and cleanups split out from the validator monitoring branch, so
as to reduce conflict area vs other PR:s
* add constants for expected message timing
* name validators after the messages they validate, mostly, to make
grepping easier
* unify field naming of EpochInfo across forks to make cross-fork code
easier
Currently, we don't have a good answer to the question "are we synced
yet" - the sync manager syncs based on the peers it's connected to, but
just because some peer looks like it should be synced from doesn't mean
we're out of sync.
Instead, we use a very silly time-based heuristic - the problem with
that is that the network can go into a rut where nobody produces blocks
- better heuristics would be needed here, but in the meantime, a command
line option can get us out of a tight spot - this PR places such an
option in the client, in the unlikely event it should be needed (most
likely in a testnet).
* ncli_db: add putState, putBlock
These tools allow modifying an existing nimbus database for the purpose
of recovery or reorg, moving the head, tail and genesis to arbitrary
points.
* remove potentially expensive `putState` in `BeaconStateDB`
* introduce `latest_block_root` which computes the root of the latest
applied block from the `latest_block_header` field (instead of passing
it in separately)
* avoid some unnecessary BeaconState copies during init
* discover https://github.com/nim-lang/Nim/issues/19094
* prefer `HashedBeaconState` in a few places to avoid recomputing state
root
* fetch latest block root from state when creating blocks
* harden `get_beacon_proposer_index` against invalid slots and document
* move random spec function tests to `test_spec.nim`
* avoid unnecessary state root computation before block proposal
* `SyncCommitteeIndex` -> `SyncSubcommitteeIndex`
* `syncCommitteePeriod` -> `sync_committee_period` (spec spelling)
* tighten period comparisons
* fix assert when validating committee message with non-altair state in
REST api
* register vc duties with subnet tracker
* fix activation logging during startup
* cache slot signature to avoid duplicate signature work
* schedule aggregation duties one slot at a time to avoid CPU spike at
each epoch
* lower aggregation subnet pre-subscription time to 4 slots (lowers
bandwidth and CPU usage)
* update stability subnets in ENR on startup
* log gossip state
* perform gossip subscriptions just before the next slot starts
* document stuff
* add random include
* don't overwrite subscription state when not subscribed
* log target gossip state
* updating gossip status once is enough
* add test
* remove syncQueueLen - this one is not updated at the end of the sync
and may cause gossip to disconnect itself completely - use a simple head
distance instead
* fix gossip disconnection - if in hysteresis, node.gossipState will be
set to disabled even though we don't disable topic subscriptions
* fix extra duty registration call
So far, `withState` and `withBlck` templates could only be used to have
convenience access to fork-agnostic BeaconState and BeaconBlock fields.
This patch:
- injects an additional `stateFork` constant that allows to use
`when` expressions to also access Altair and Merge-specific fields.
- introduces a `withStateAndBlck` template to support operating on both
a `BeaconState` and `BeaconBlock` at a time.
- makes sync committee related functions Merge aware.
- changes a couple if-else trees for forks into case statements so that
forgotten future forks are promoted to compile-time errors.
There are a number of locations in the code that get attestations on a
forked beacon state. For attestation pools test, a convenience wrapper
was available to reduce clutter. This patch integrates that wrapper into
the core component so that it can also take advantage of the wrapper.
The initialization of a `SyncAggregate` to its default value is not very
intuitive. There is an `init` function in `sync_committee_msg_pool` that
provides a convenience wrapper. This patch exports that initializer so
that the rest of the code base can also take advantage of it.
The phase0 and altair overloads of `makeBeaconBlock` slightly differ in
their signatures which makes using them unnecessarily verbose.
- A placeholder `sync_aggregate` argument similar to `executionPayload`
is added to the phase0 overload to match the altair signature.
- A wrapper operating on `ForkedHashedBeaconState` is introduced.
* Placing callbacks into strategic places.
* Initial events call implementation.
* Post rebase fixes.
* Change addSyncContribution() implementation.
* Add `attestation-sent` event.
Remove gcsafe, raises from callbacks implementations.
Move `attestation-received` fire at the end of attestation processing.
* Address review comments.
Jacek Sieka
tersec
Etan Kissling
zah
Mamy Ratsimbazafy
Eugene Kabanov