## Issue Addressed
Closes#4481.
(Continuation of #4648)
## Proposed Changes
- [x] Add `lighthouse db prune-states`
- [x] Make it work
- [x] Ensure block roots are handled correctly (to be addressed in 4735)
- [x] Check perf on mainnet/Goerli/Gnosis (takes a few seconds max)
- [x] Run block root healing logic (#4875 ) at the beginning
- [x] Add some tests
- [x] Update docs
- [x] Add `--freezer` flag and other improvements to `lighthouse db inspect`
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
Co-authored-by: Jimmy Chen <jimmy@sigmaprime.io>
Co-authored-by: Michael Sproul <micsproul@gmail.com>
## Issue Addressed
Closes#4817.
## Proposed Changes
- Fill in the linear block roots array between 0 and the slot of the first block (e.g. slots 0 and 1 on Holesky).
- Backport the `--freezer`, `--skip` and `--limit` options for `lighthouse db inspect` from tree-states. This allows us to easily view the database corruption of 4817 using `lighthouse db inspect --network holesky --freezer --column bbr --output values --limit 2`.
- Backport the `iter_column_from` change and `MemoryStore` overhaul from tree-states. These are required to enable `lighthouse db inspect`.
- Rework `freezer_upper_limit` to allow state lookups for slots below the `state_lower_limit`. Currently state lookups will fail until state reconstruction completes entirely.
There is a new regression test for the main bug, but no test for the `freezer_upper_limit` fix because we don't currently support running state reconstruction partially (see #3026). This will be fixed once we merge `tree-states`! In lieu of an automated test, I've tested manually on a Holesky node while it was reconstructing.
## Additional Info
Users who backfilled Holesky to slot 0 (e.g. using `--reconstruct-historic-states`) need to either:
- Re-sync from genesis.
- Re-sync using checkpoint sync and the changes from this PR.
Due to the recency of the Holesky genesis, writing a custom pass to fix up broken databases (which would require its own thorough testing) was deemed unnecessary. This is the primary reason for this PR being marked `backwards-incompat`.
This will create few conflicts with Deneb, which I've already resolved on `tree-states-deneb` and will be happy to backport to Deneb once this PR is merged to unstable.
## Issue Addressed
Fixes#4697.
This also unblocks the state pruning PR (#4835). Because self healing breaks if state pruning is applied to a database with missing block roots.
## Proposed Changes
- Fill in the missing block roots between last restore point slot and split slot when upgrading to latest database version.
## Issue Addressed
Makes lighthouse compliant with new kzg changes in https://github.com/ethereum/consensus-specs/releases/tag/v1.4.0-beta.3
## Proposed Changes
1. Adds new official trusted setup
2. Refactors kzg to match upstream changes in https://github.com/ethereum/c-kzg-4844/pull/377
3. Updates pre-generated `BlobBundle` to work with official trusted setup. ~~Using json here instead of ssz to account for different value of `MaxBlobCommitmentsPerBlock` in minimal and mainnet. By using json, we can just use one pre generated bundle for both minimal and mainnet. Size of 2 separate ssz bundles is approximately equal to one json bundle cc @jimmygchen~~
Dunno what I was doing, ssz works without any issues
4. Stores trusted_setup as just bytes in eth2_network_config so that we don't have kzg dependency in that lib and in lcli.
Co-authored-by: realbigsean <seananderson33@gmail.com>
Co-authored-by: realbigsean <seananderson33@GMAIL.com>
* add processing and processed caching to the DA checker
* move processing cache out of critical cache
* get it compiling
* fix lints
* add docs to `AvailabilityView`
* some self review
* fix lints
* fix beacon chain tests
* cargo fmt
* make availability view easier to implement, start on testing
* move child component cache and finish test
* cargo fix
* cargo fix
* cargo fix
* fmt and lint
* make blob commitments not optional, rename some caches, add missing blobs struct
* Update beacon_node/beacon_chain/src/data_availability_checker/processing_cache.rs
Co-authored-by: ethDreamer <37123614+ethDreamer@users.noreply.github.com>
* marks review feedback and other general cleanup
* cargo fix
* improve availability view docs
* some renames
* some renames and docs
* fix should delay lookup logic
* get rid of some wrapper methods
* fix up single lookup changes
* add a couple docs
* add single blob merge method and improve process_... docs
* update some names
* lints
* fix merge
* remove blob indices from lookup creation log
* remove blob indices from lookup creation log
* delayed lookup logging improvement
* check fork choice before doing any blob processing
* remove unused dep
* Update beacon_node/beacon_chain/src/data_availability_checker/availability_view.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Update beacon_node/beacon_chain/src/data_availability_checker/availability_view.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Update beacon_node/beacon_chain/src/data_availability_checker/availability_view.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Update beacon_node/beacon_chain/src/data_availability_checker/availability_view.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Update beacon_node/network/src/sync/block_lookups/delayed_lookup.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* remove duplicate deps
* use gen range in random blobs geneartor
* rename processing cache fields
* require block root in rpc block construction and check block root consistency
* send peers as vec in single message
* spawn delayed lookup service from network beacon processor
* fix tests
---------
Co-authored-by: ethDreamer <37123614+ethDreamer@users.noreply.github.com>
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Start testing blob pruning
* Get rid of unnecessary orphaned blob column
* Make random blob tests deterministic
* Test for pruning being blocked by finality
* Fix bugs and test fork boundary
* A few more tweaks to pruning conditions
* Tweak oldest_blob_slot semantics
* Test margin pruning
* Clean up some terminology and lints
* Schema migrations for v18
* Remove FIXME
* Prune blobs on finalization not every slot
* Fix more bugs + tests
* Address review comments
## Issue Addressed
Fix a bug in the storage of the linear block roots array in the freezer DB. Previously this array was always written as part of state storage (or block backfill). With state pruning enabled by #4610, these states were no longer being written and as a result neither were the block roots.
The impact is quite low, we would just log an error when trying to forwards-iterate the block roots, which for validating nodes only happens when they try to look up blocks for peers:
> Aug 25 03:42:36.980 ERRO Missing chunk in forwards iterator chunk index: 49726, service: freezer_db
Any node checkpoint synced off `unstable` is affected and has a corrupt database. If you see the log above, you need to re-sync with the fix. Nodes that haven't checkpoint synced recently should _not_ be corrupted, even if they ran the buggy version.
## Proposed Changes
- Use a `ChunkWriter` to write the block roots when states are not being stored.
- Tweak the usage of `get_latest_restore_point` so that it doesn't return a nonsense value when state pruning is enabled.
- Tweak the guarantee on the block roots array so that block roots are assumed available up to the split slot (exclusive). This is a bit nicer than relying on anything to do with the latest restore point, which is a nonsensical concept when there aren't any restore points.
## Additional Info
I'm looking forward to deleting the chunked vector code for good when we merge tree-states 😁
## Issue Addressed
Closes#3210Closes#3211
## Proposed Changes
- Checkpoint sync from the latest finalized state regardless of its alignment.
- Add the `block_root` to the database's split point. This is _only_ added to the in-memory split in order to avoid a schema migration. See `load_split`.
- Add a new method to the DB called `get_advanced_state`, which looks up a state _by block root_, with a `state_root` as fallback. Using this method prevents accidental accesses of the split's unadvanced state, which does not exist in the hot DB and is not guaranteed to exist in the freezer DB at all. Previously Lighthouse would look up this state _from the freezer DB_, even if it was required for block/attestation processing, which was suboptimal.
- Replace several state look-ups in block and attestation processing with `get_advanced_state` so that they can't hit the split block's unadvanced state.
- Do not store any states in the freezer database by default. All states will be deleted upon being evicted from the hot database unless `--reconstruct-historic-states` is set. The anchor info which was previously used for checkpoint sync is used to implement this, including when syncing from genesis.
## Additional Info
Needs further testing. I want to stress-test the pruned database under Hydra.
The `get_advanced_state` method is intended to become more relevant over time: `tree-states` includes an identically named method that returns advanced states from its in-memory cache.
Co-authored-by: realbigsean <seananderson33@gmail.com>
* some blob reprocessing work
* remove ForceBlockLookup
* reorder enum match arms in sync manager
* a lot more reprocessing work
* impl logic for triggerng blob lookups along with block lookups
* deal with rpc blobs in groups per block in the da checker. don't cache missing blob ids in the da checker.
* make single block lookup generic
* more work
* add delayed processing logic and combine some requests
* start fixing some compile errors
* fix compilation in main block lookup mod
* much work
* get things compiling
* parent blob lookups
* fix compile
* revert red/stevie changes
* fix up sync manager delay message logic
* add peer usefulness enum
* should remove lookup refactor
* consolidate retry error handling
* improve peer scoring during certain failures in parent lookups
* improve retry code
* drop parent lookup if either req has a peer disconnect during download
* refactor single block processed method
* processing peer refactor
* smol bugfix
* fix some todos
* fix lints
* fix lints
* fix compile in lookup tests
* fix lints
* fix lints
* fix existing block lookup tests
* renamings
* fix after merge
* cargo fmt
* compilation fix in beacon chain tests
* fix
* refactor lookup tests to work with multiple forks and response types
* make tests into macros
* wrap availability check error
* fix compile after merge
* add random blobs
* start fixing up lookup verify error handling
* some bug fixes and the start of deneb only tests
* make tests work for all forks
* track information about peer source
* error refactoring
* improve peer scoring
* fix test compilation
* make sure blobs are sent for processing after stream termination, delete copied tests
* add some tests and fix a bug
* smol bugfixes and moar tests
* add tests and fix some things
* compile after merge
* lots of refactoring
* retry on invalid block/blob
* merge unknown parent messages before current slot lookup
* get tests compiling
* penalize blob peer on invalid blobs
* Check disk on in-memory cache miss
* Update beacon_node/beacon_chain/src/data_availability_checker/overflow_lru_cache.rs
* Update beacon_node/network/src/sync/network_context.rs
Co-authored-by: Divma <26765164+divagant-martian@users.noreply.github.com>
* fix bug in matching blocks and blobs in range sync
* pr feedback
* fix conflicts
* upgrade logs from warn to crit when we receive incorrect response in range
* synced_and_connected_within_tolerance -> should_search_for_block
* remove todo
* add data gas used and update excess data gas to u64
* Fix Broken Overflow Tests
* payload verification with commitments
* fix merge conflicts
* restore payload file
* Restore payload file
* remove todo
* add max blob commitments per block
* c-kzg lib update
* Fix ef tests
* Abstract over minimal/mainnet spec in kzg crate
* Start integrating new KZG
* checkpoint sync without alignment
* checkpoint sync without alignment
* add import
* add import
* query for checkpoint state by slot rather than state root (teku doesn't serve by state root)
* query for checkpoint state by slot rather than state root (teku doesn't serve by state root)
* loosen check
* get state first and query by most recent block root
* Revert "loosen check"
This reverts commit 069d13dd63aa794a3505db9f17bd1a6b73f0be81.
* get state first and query by most recent block root
* merge max blobs change
* simplify delay logic
* rename unknown parent sync message variants
* rename parameter, block_slot -> slot
* add some docs to the lookup module
* use interval instead of sleep
* drop request if blocks and blobs requests both return `None` for `Id`
* clean up `find_single_lookup` logic
* add lookup source enum
* clean up `find_single_lookup` logic
* add docs to find_single_lookup_request
* move LookupSource our of param where unnecessary
* remove unnecessary todo
* query for block by `state.latest_block_header.slot`
* fix lint
* fix merge transition ef tests
* fix test
* fix test
* fix observed blob sidecars test
* Add some metrics (#33)
* fix protocol limits for blobs by root
* Update Engine API for 1:1 Structure Method
* make beacon chain tests to fix devnet 6 changes
* get ckzg working and fix some tests
* fix remaining tests
* fix lints
* Fix KZG linking issues
* remove unused dep
* lockfile
* test fixes
* remove dbgs
* remove unwrap
* cleanup tx generator
* small fixes
* fixing fixes
* more self reivew
* more self review
* refactor genesis header initialization
* refactor mock el instantiations
* fix compile
* fix network test, make sure they run for each fork
* pr feedback
* fix last test (hopefully)
---------
Co-authored-by: Pawan Dhananjay <pawandhananjay@gmail.com>
Co-authored-by: Mark Mackey <mark@sigmaprime.io>
Co-authored-by: Divma <26765164+divagant-martian@users.noreply.github.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
## Issue Addressed
- #4293
- #4264
## Proposed Changes
*Changes largely follow those suggested in the main issue*.
- Add new routes to HTTP API
- `post_beacon_blocks_v2`
- `post_blinded_beacon_blocks_v2`
- Add new routes to `BeaconNodeHttpClient`
- `post_beacon_blocks_v2`
- `post_blinded_beacon_blocks_v2`
- Define new Eth2 common types
- `BroadcastValidation`, enum representing the level of validation to apply to blocks prior to broadcast
- `BroadcastValidationQuery`, the corresponding HTTP query string type for the above type
- ~~Define `_checked` variants of both `publish_block` and `publish_blinded_block` that enforce a validation level at a type level~~
- Add interactive tests to the `bn_http_api_tests` test target covering each validation level (to their own test module, `broadcast_validation_tests`)
- `beacon/blocks`
- `broadcast_validation=gossip`
- Invalid (400)
- Full Pass (200)
- Partial Pass (202)
- `broadcast_validation=consensus`
- Invalid (400)
- Only gossip (400)
- Only consensus pass (i.e., equivocates) (200)
- Full pass (200)
- `broadcast_validation=consensus_and_equivocation`
- Invalid (400)
- Invalid due to early equivocation (400)
- Only gossip (400)
- Only consensus (400)
- Pass (200)
- `beacon/blinded_blocks`
- `broadcast_validation=gossip`
- Invalid (400)
- Full Pass (200)
- Partial Pass (202)
- `broadcast_validation=consensus`
- Invalid (400)
- Only gossip (400)
- ~~Only consensus pass (i.e., equivocates) (200)~~
- Full pass (200)
- `broadcast_validation=consensus_and_equivocation`
- Invalid (400)
- Invalid due to early equivocation (400)
- Only gossip (400)
- Only consensus (400)
- Pass (200)
- Add a new trait, `IntoGossipVerifiedBlock`, which allows type-level guarantees to be made as to gossip validity
- Modify the structure of the `ObservedBlockProducers` cache from a `(slot, validator_index)` mapping to a `((slot, validator_index), block_root)` mapping
- Modify `ObservedBlockProducers::proposer_has_been_observed` to return a `SeenBlock` rather than a boolean on success
- Punish gossip peer (low) for submitting equivocating blocks
- Rename `BlockError::SlashablePublish` to `BlockError::SlashableProposal`
## Additional Info
This PR contains changes that directly modify how blocks are verified within the client. For more context, consult [comments in-thread](https://github.com/sigp/lighthouse/pull/4316#discussion_r1234724202).
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
## Issue Addressed
Closes#4332
## Proposed Changes
Remove the `CountUnrealized` type, defaulting unrealized justification to _on_. This fixes the #4332 issue by ensuring that importing the same block to fork choice always results in the same outcome.
Finalized sync speed may be slightly impacted by this change, but that is deemed an acceptable trade-off until the optimisation from #4118 is implemented.
TODO:
- [x] Also check that the block isn't a duplicate before importing
* Update Engine API to Latest
* Get Mock EE Working
* Fix Mock EE
* Update Engine API Again
* Rip out get_blobs_bundle Stuff
* Fix Test Harness
* Fix Clippy Complaints
* Fix Beacon Chain Tests
* Use eth1_withdrawal_credential in Some Test States
* Update beacon_node/genesis/src/interop.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Update beacon_node/genesis/src/interop.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Increase validator sizes
* Pick next sync committee message
Co-authored-by: Michael Sproul <micsproul@gmail.com>
Co-authored-by: Paul Hauner <paul@paulhauner.com>
* Use eth1_withdrawal_credential in Some Test States
* Update beacon_node/genesis/src/interop.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Update beacon_node/genesis/src/interop.rs
Co-authored-by: Michael Sproul <micsproul@gmail.com>
* Increase validator sizes
* Pick next sync committee message
Co-authored-by: Michael Sproul <micsproul@gmail.com>
Co-authored-by: Paul Hauner <paul@paulhauner.com>
## Issue Addressed
NA
## Proposed Changes
Myself and others (#3678) have observed that when running with lots of validators (e.g., 1000s) the cardinality is too much for Prometheus. I've seen Prometheus instances just grind to a halt when we turn the validator monitor on for our testnet validators (we have 10,000s of Goerli validators). Additionally, the debug log volume can get very high with one log per validator, per attestation.
To address this, the `bn --validator-monitor-individual-tracking-threshold <INTEGER>` flag has been added to *disable* per-validator (i.e., non-aggregated) metrics/logging once the validator monitor exceeds the threshold of validators. The default value is `64`, which is a finger-to-the-wind value. I don't actually know the value at which Prometheus starts to become overwhelmed, but I've seen it work with ~64 validators and I've seen it *not* work with 1000s of validators. A default of `64` seems like it will result in a breaking change to users who are running millions of dollars worth of validators whilst resulting in a no-op for low-validator-count users. I'm open to changing this number, though.
Additionally, this PR starts collecting aggregated Prometheus metrics (e.g., total count of head hits across all validators), so that high-validator-count validators still have some interesting metrics. We already had logging for aggregated values, so nothing has been added there.
I've opted to make this a breaking change since it can be rather damaging to your Prometheus instance to accidentally enable the validator monitor with large numbers of validators. I've crashed a Prometheus instance myself and had a report from another user who's done the same thing.
## Additional Info
NA
## Breaking Changes Note
A new label has been added to the validator monitor Prometheus metrics: `total`. This label tracks the aggregated metrics of all validators in the validator monitor (as opposed to each validator being tracking individually using its pubkey as the label).
Additionally, a new flag has been added to the Beacon Node: `--validator-monitor-individual-tracking-threshold`. The default value is `64`, which means that when the validator monitor is tracking more than 64 validators then it will stop tracking per-validator metrics and only track the `all_validators` metric. It will also stop logging per-validator logs and only emit aggregated logs (the exception being that exit and slashing logs are always emitted).
These changes were introduced in #3728 to address issues with untenable Prometheus cardinality and log volume when using the validator monitor with high validator counts (e.g., 1000s of validators). Users with less than 65 validators will see no change in behavior (apart from the added `all_validators` metric). Users with more than 65 validators who wish to maintain the previous behavior can set something like `--validator-monitor-individual-tracking-threshold 999999`.
## Issue Addressed
#3704
## Proposed Changes
Adds is_syncing_finalized: bool parameter for block verification functions. Sets the payload_verification_status to Optimistic if is_syncing_finalized is true. Uses SyncState in NetworkGlobals in BeaconProcessor to retrieve the syncing status.
## Additional Info
I could implement FinalizedSignatureVerifiedBlock if you think it would be nicer.
## Issue Addressed
Fix a bug in block production that results in blocks with 0 attestations during the first slot of an epoch.
The bug is marked by debug logs of the form:
> DEBG Discarding attestation because of missing ancestor, block_root: 0x3cc00d9c9e0883b2d0db8606278f2b8423d4902f9a1ee619258b5b60590e64f8, pivot_slot: 4042591
It occurs when trying to look up the shuffling decision root for an attestation from a slot which is prior to fork choice's finalized block. This happens frequently when proposing in the first slot of the epoch where we have:
- `current_epoch == n`
- `attestation.data.target.epoch == n - 1`
- attestation shuffling epoch `== n - 3` (decision block being the last block of `n - 3`)
- `state.finalized_checkpoint.epoch == n - 2` (first block of `n - 2` is finalized)
Hence the shuffling decision slot is out of range of the fork choice backwards iterator _by a single slot_.
Unfortunately this bug was hidden when we weren't pruning fork choice, and then reintroduced in v2.5.1 when we fixed the pruning (https://github.com/sigp/lighthouse/releases/tag/v2.5.1). There's no way to turn that off or disable the filtering in our current release, so we need a new release to fix this issue.
Fortunately, it also does not occur on every epoch boundary because of the gradual pruning of fork choice every 256 blocks (~8 epochs):
01e84b71f5/consensus/proto_array/src/proto_array_fork_choice.rs (L16)01e84b71f5/consensus/proto_array/src/proto_array.rs (L713-L716)
So the probability of proposing a 0-attestation block given a proposal assignment is approximately `1/32 * 1/8 = 0.39%`.
## Proposed Changes
- Load the block's shuffling ID from fork choice and verify it against the expected shuffling ID of the head state. This code was initially written before we had settled on a representation of shuffling IDs, so I think it's a nice simplification to make use of them here rather than more ad-hoc logic that fundamentally does the same thing.
## Additional Info
Thanks to @moshe-blox for noticing this issue and bringing it to our attention.
## Issue Addressed
NA
## Proposed Changes
This PR attempts to fix the following spurious CI failure:
```
---- store_tests::garbage_collect_temp_states_from_failed_block stdout ----
thread 'store_tests::garbage_collect_temp_states_from_failed_block' panicked at 'disk store should initialize: DBError { message: "Error { message: \"IO error: lock /tmp/.tmp6DcBQ9/cold_db/LOCK: already held by process\" }" }', beacon_node/beacon_chain/tests/store_tests.rs:59:10
```
I believe that some async task is taking a clone of the store and holding it in some other thread for a short time. This creates a race-condition when we try to open a new instance of the store.
## Additional Info
NA
## Issue Addressed
NA
## Proposed Changes
This PR removes duplicated block root computation.
Computing the `SignedBeaconBlock::canonical_root` has become more expensive since the merge as we need to compute the merke root of each transaction inside an `ExecutionPayload`.
Computing the root for [a mainnet block](https://beaconcha.in/slot/4704236) is taking ~10ms on my i7-8700K CPU @ 3.70GHz (no sha extensions). Given that our median seen-to-imported time for blocks is presently 300-400ms, removing a few duplicated block roots (~30ms) could represent an easy 10% improvement. When we consider that the seen-to-imported times include operations *after* the block has been placed in the early attester cache, we could expect the 30ms to be more significant WRT our seen-to-attestable times.
## Additional Info
NA
## Issue Addressed
Closes https://github.com/sigp/lighthouse/issues/3556
## Proposed Changes
Delete finalized execution payloads from the database in two places:
1. When running the finalization migration in `migrate_database`. We delete the finalized payloads between the last split point and the new updated split point. _If_ payloads are already pruned prior to this then this is sufficient to prune _all_ payloads as non-canonical payloads are already deleted by the head pruner, and all canonical payloads prior to the previous split will already have been pruned.
2. To address the fact that users will update to this code _after_ the merge on mainnet (and testnets), we need a one-off scan to delete the finalized payloads from the canonical chain. This is implemented in `try_prune_execution_payloads` which runs on startup and scans the chain back to the Bellatrix fork or the anchor slot (if checkpoint synced after Bellatrix). In the case where payloads are already pruned this check only imposes a single state load for the split state, which shouldn't be _too slow_. Even so, a flag `--prepare-payloads-on-startup=false` is provided to turn this off after it has run the first time, which provides faster start-up times.
There is also a new `lighthouse db prune_payloads` subcommand for users who prefer to run the pruning manually.
## Additional Info
The tests have been updated to not rely on finalized payloads in the database, instead using the `MockExecutionLayer` to reconstruct them. Additionally a check was added to `check_chain_dump` which asserts the non-existence or existence of payloads on disk depending on their slot.
## Proposed Changes
This PR has two aims: to speed up attestation packing in the op pool, and to fix bugs in the verification of attester slashings, proposer slashings and voluntary exits. The changes are bundled into a single database schema upgrade (v12).
Attestation packing is sped up by removing several inefficiencies:
- No more recalculation of `attesting_indices` during packing.
- No (unnecessary) examination of the `ParticipationFlags`: a bitfield suffices. See `RewardCache`.
- No re-checking of attestation validity during packing: the `AttestationMap` provides attestations which are "correct by construction" (I have checked this using Hydra).
- No SSZ re-serialization for the clunky `AttestationId` type (it can be removed in a future release).
So far the speed-up seems to be roughly 2-10x, from 500ms down to 50-100ms.
Verification of attester slashings, proposer slashings and voluntary exits is fixed by:
- Tracking the `ForkVersion`s that were used to verify each message inside the `SigVerifiedOp`. This allows us to quickly re-verify that they match the head state's opinion of what the `ForkVersion` should be at the epoch(s) relevant to the message.
- Storing the `SigVerifiedOp` on disk rather than the raw operation. This allows us to continue track the fork versions after a reboot.
This is mostly contained in this commit 52bb1840ae5c4356a8fc3a51e5df23ed65ed2c7f.
## Additional Info
The schema upgrade uses the justified state to re-verify attestations and compute `attesting_indices` for them. It will drop any attestations that fail to verify, by the logic that attestations are most valuable in the few slots after they're observed, and are probably stale and useless by the time a node restarts. Exits and proposer slashings and similarly re-verified to obtain `SigVerifiedOp`s.
This PR contains a runtime killswitch `--paranoid-block-proposal` which opts out of all the optimisations in favour of closely verifying every included message. Although I'm quite sure that the optimisations are correct this flag could be useful in the event of an unforeseen emergency.
Finally, you might notice that the `RewardCache` appears quite useless in its current form because it is only updated on the hot-path immediately before proposal. My hope is that in future we can shift calls to `RewardCache::update` into the background, e.g. while performing the state advance. It is also forward-looking to `tree-states` compatibility, where iterating and indexing `state.{previous,current}_epoch_participation` is expensive and needs to be minimised.
## Issue Addressed
NA
## Proposed Changes
There are scenarios where the only viable head will have an invalid execution payload, in this scenario the `get_head` function on `proto_array` will return an error. We must recover from this scenario by importing blocks from the network.
This PR stops `BeaconChain::recompute_head` from returning an error so that we can't accidentally start down-scoring peers or aborting block import just because the current head has an invalid payload.
## Reviewer Notes
The following changes are included:
1. Allow `fork_choice.get_head` to fail gracefully in `BeaconChain::process_block` when trying to update the `early_attester_cache`; simply don't add the block to the cache rather than aborting the entire process.
1. Don't return an error from `BeaconChain::recompute_head_at_current_slot` and `BeaconChain::recompute_head` to defensively prevent calling functions from aborting any process just because the fork choice function failed to run.
- This should have practically no effect, since most callers were still continuing if recomputing the head failed.
- The outlier is that the API will return 200 rather than a 500 when fork choice fails.
1. Add the `ProtoArrayForkChoice::set_all_blocks_to_optimistic` function to recover from the scenario where we've rebooted and the persisted fork choice has an invalid head.
## Issue Addressed
Add a flag that optionally enables unrealized vote tracking. Would like to test out on testnets and benchmark differences in methods of vote tracking. This PR includes a DB schema upgrade to enable to new vote tracking style.
Co-authored-by: realbigsean <sean@sigmaprime.io>
Co-authored-by: Paul Hauner <paul@paulhauner.com>
Co-authored-by: sean <seananderson33@gmail.com>
Co-authored-by: Mac L <mjladson@pm.me>
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough 😅)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](de2b2801c8/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java (L171-L182)) it uses [`getStateRootFromBlockRoot`](de2b2801c8/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java (L336-L341)) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
## Proposed Changes
Reduce post-merge disk usage by not storing finalized execution payloads in Lighthouse's database.
⚠️ **This is achieved in a backwards-incompatible way for networks that have already merged** ⚠️. Kiln users and shadow fork enjoyers will be unable to downgrade after running the code from this PR. The upgrade migration may take several minutes to run, and can't be aborted after it begins.
The main changes are:
- New column in the database called `ExecPayload`, keyed by beacon block root.
- The `BeaconBlock` column now stores blinded blocks only.
- Lots of places that previously used full blocks now use blinded blocks, e.g. analytics APIs, block replay in the DB, etc.
- On finalization:
- `prune_abanonded_forks` deletes non-canonical payloads whilst deleting non-canonical blocks.
- `migrate_db` deletes finalized canonical payloads whilst deleting finalized states.
- Conversions between blinded and full blocks are implemented in a compositional way, duplicating some work from Sean's PR #3134.
- The execution layer has a new `get_payload_by_block_hash` method that reconstructs a payload using the EE's `eth_getBlockByHash` call.
- I've tested manually that it works on Kiln, using Geth and Nethermind.
- This isn't necessarily the most efficient method, and new engine APIs are being discussed to improve this: https://github.com/ethereum/execution-apis/pull/146.
- We're depending on the `ethers` master branch, due to lots of recent changes. We're also using a workaround for https://github.com/gakonst/ethers-rs/issues/1134.
- Payload reconstruction is used in the HTTP API via `BeaconChain::get_block`, which is now `async`. Due to the `async` fn, the `blocking_json` wrapper has been removed.
- Payload reconstruction is used in network RPC to serve blocks-by-{root,range} responses. Here the `async` adjustment is messier, although I think I've managed to come up with a reasonable compromise: the handlers take the `SendOnDrop` by value so that they can drop it on _task completion_ (after the `fn` returns). Still, this is introducing disk reads onto core executor threads, which may have a negative performance impact (thoughts appreciated).
## Additional Info
- [x] For performance it would be great to remove the cloning of full blocks when converting them to blinded blocks to write to disk. I'm going to experiment with a `put_block` API that takes the block by value, breaks it into a blinded block and a payload, stores the blinded block, and then re-assembles the full block for the caller.
- [x] We should measure the latency of blocks-by-root and blocks-by-range responses.
- [x] We should add integration tests that stress the payload reconstruction (basic tests done, issue for more extensive tests: https://github.com/sigp/lighthouse/issues/3159)
- [x] We should (manually) test the schema v9 migration from several prior versions, particularly as blocks have changed on disk and some migrations rely on being able to load blocks.
Co-authored-by: Paul Hauner <paul@paulhauner.com>
## Issue Addressed
Fix a panic due to misuse of the Tokio executor when processing a forkchoiceUpdated response. We were previously calling `process_invalid_execution_payload` from the async function `update_execution_engine_forkchoice_async`, which resulted in a panic because `process_invalid_execution_payload` contains a call to fork choice, which ultimately calls `block_on`.
An example backtrace can be found here: https://gist.github.com/michaelsproul/ac5da03e203d6ffac672423eaf52fb20
## Proposed Changes
Wrap the call to `process_invalid_execution_payload` in a `spawn_blocking` so that `block_on` is no longer called from an async context.
## Additional Info
- I've been thinking about how to catch bugs like this with static analysis (a new Clippy lint).
- The payload validation tests have been re-worked to support distinct responses from the mock EE for newPayload and forkchoiceUpdated. Three new tests have been added covering the `Invalid`, `InvalidBlockHash` and `InvalidTerminalBlock` cases.
- I think we need a bunch more tests of different legal and illegal variations
## Issue Addressed
NA
## Proposed Changes
Adds the functionality to allow blocks to be validated/invalidated after their import as per the [optimistic sync spec](https://github.com/ethereum/consensus-specs/blob/dev/sync/optimistic.md#how-to-optimistically-import-blocks). This means:
- Updating `ProtoArray` to allow flipping the `execution_status` of ancestors/descendants based on payload validity updates.
- Creating separation between `execution_layer` and the `beacon_chain` by creating a `PayloadStatus` struct.
- Refactoring how the `execution_layer` selects a `PayloadStatus` from the multiple statuses returned from multiple EEs.
- Adding testing framework for optimistic imports.
- Add `ExecutionBlockHash(Hash256)` new-type struct to avoid confusion between *beacon block roots* and *execution payload hashes*.
- Add `merge` to [`FORKS`](c3a793fd73/Makefile (L17)) in the `Makefile` to ensure we test the beacon chain with merge settings.
- Fix some tests here that were failing due to a missing execution layer.
## TODO
- [ ] Balance tests
Co-authored-by: Mark Mackey <mark@sigmaprime.io>
## Issue Addressed
Successor to #2431
## Proposed Changes
* Add a `BlockReplayer` struct to abstract over the intricacies of calling `per_slot_processing` and `per_block_processing` while avoiding unnecessary tree hashing.
* Add a variant of the forwards state root iterator that does not require an `end_state`.
* Use the `BlockReplayer` when reconstructing states in the database. Use the efficient forwards iterator for frozen states.
* Refactor the iterators to remove `Arc<HotColdDB>` (this seems to be neater than making _everything_ an `Arc<HotColdDB>` as I did in #2431).
Supplying the state roots allow us to avoid building a tree hash cache at all when reconstructing historic states, which saves around 1 second flat (regardless of `slots-per-restore-point`). This is a small percentage of worst-case state load times with 200K validators and SPRP=2048 (~15s vs ~16s) but a significant speed-up for more frequent restore points: state loads with SPRP=32 should be now consistently <500ms instead of 1.5s (a ~3x speedup).
## Additional Info
Required by https://github.com/sigp/lighthouse/pull/2628
## Issue Addressed
Resolves: https://github.com/sigp/lighthouse/issues/2741
Includes: https://github.com/sigp/lighthouse/pull/2853 so that we can get ssz static tests passing here on v1.1.6. If we want to merge that first, we can make this diff slightly smaller
## Proposed Changes
- Changes the `justified_epoch` and `finalized_epoch` in the `ProtoArrayNode` each to an `Option<Checkpoint>`. The `Option` is necessary only for the migration, so not ideal. But does allow us to add a default logic to `None` on these fields during the database migration.
- Adds a database migration from a legacy fork choice struct to the new one, search for all necessary block roots in fork choice by iterating through blocks in the db.
- updates related to https://github.com/ethereum/consensus-specs/pull/2727
- We will have to update the persisted forkchoice to make sure the justified checkpoint stored is correct according to the updated fork choice logic. This boils down to setting the forkchoice store's justified checkpoint to the justified checkpoint of the block that advanced the finalized checkpoint to the current one.
- AFAICT there's no migration steps necessary for the update to allow applying attestations from prior blocks, but would appreciate confirmation on that
- I updated the consensus spec tests to v1.1.6 here, but they will fail until we also implement the proposer score boost updates. I confirmed that the previously failing scenario `new_finalized_slot_is_justified_checkpoint_ancestor` will now pass after the boost updates, but haven't confirmed _all_ tests will pass because I just quickly stubbed out the proposer boost test scenario formatting.
- This PR now also includes proposer boosting https://github.com/ethereum/consensus-specs/pull/2730
## Additional Info
I realized checking justified and finalized roots in fork choice makes it more likely that we trigger this bug: https://github.com/ethereum/consensus-specs/pull/2727
It's possible the combination of justified checkpoint and finalized checkpoint in the forkchoice store is different from in any block in fork choice. So when trying to startup our store's justified checkpoint seems invalid to the rest of fork choice (but it should be valid). When this happens we get an `InvalidBestNode` error and fail to start up. So I'm including that bugfix in this branch.
Todo:
- [x] Fix fork choice tests
- [x] Self review
- [x] Add fix for https://github.com/ethereum/consensus-specs/pull/2727
- [x] Rebase onto Kintusgi
- [x] Fix `num_active_validators` calculation as @michaelsproul pointed out
- [x] Clean up db migrations
Co-authored-by: realbigsean <seananderson33@gmail.com>
