#7727 introduced a bug in the logging, where as long as the node failed the SSZ `get_validator_blocks_v3` endpoint, it would log as `Beacon node does not support...`. However, the failure can be due to other reasons, such as a timed out error as found by @jimmygchen:
`WARN Beacon node does not support SSZ in block production, falling back to JSON slot: 5283379, error: HttpClient(url: https://ho-h-bn-cowl.spesi.io:15052/, kind: timeout, detail: operation timed out`
This PR made the error log more generic, so there is less confusion.
Additionally, suggested by @michaelsproul, this PR refactors the `get_validator_blocks_v3` calls by trying all beacon nodes using the SSZ endpoint first, and if all beacon node fails the SSZ endpoint, only then fallback to JSON.
It changes the logic from:
"SSZ -> JSON for primary beacon node, followed by SSZ -> JSON for second beacon node and so on" to
"SSZ for all beacon nodes -> JSON for all beacon nodes"
This has the advantage that if the primary beacon node is having issues and failed the SSZ, we avoid retrying the primary beacon node again on JSON (as it could be that the primary beacon node fail again); rather, we switch to the second beacon node.
Co-Authored-By: Tan Chee Keong <tanck@sigmaprime.io>
Co-Authored-By: chonghe <44791194+chong-he@users.noreply.github.com>
Instrument beacon node startup and parallelise pubkey cache initialisation.
I instrumented beacon node startup and noticed that pubkey cache takes a long time to initialise, mostly due to decompressing all the validator pubkeys.
This PR uses rayon to parallelize the decompression on initial checkpoint sync. The pubkeys are stored uncompressed, so the decopression time is not a problem on subsequent restarts. On restarts, we still deserialize pubkeys, but the timing is quite minimal on Sepolia so I didn't investigate further.
`validator_pubkey_cache_new` timing on Sepolia:
* before: 109.64ms
* with parallelization: 21ms
on Hoodi:
* before: times out with Kurtosis after 120s
* with parallelization: 12.77s to import keys
**UPDATE**: downloading checkpoint state + genesis state takes about 2 minutes on my laptop, so it seems like the BN managed to start the http server just before timing out (after the optimisation).
<img width="1380" height="625" alt="image" src="https://github.com/user-attachments/assets/4c548c14-57dd-4b47-af9a-115b15791940" />
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
This pull request introduces workflows and updates to ensure reproducible builds for the Lighthouse project. It adds two GitHub Actions workflows for building and testing reproducible Docker images and binaries, updates the `Makefile` to streamline reproducible build configurations, and modifies the `Dockerfile.reproducible` to align with the new build process. Additionally, it removes the `reproducible` profile from `Cargo.toml`.
### New GitHub Actions Workflows:
* [`.github/workflows/docker-reproducible.yml`](diffhunk://#diff-222af23bee616920b04f5b92a83eb5106fce08abd885cd3a3b15b8beb5e789c3R1-R145): Adds a workflow to build and push reproducible multi-architecture Docker images for releases, including support for dry runs without pushing an image.
### Build Configuration Updates:
* [`Makefile`](diffhunk://#diff-76ed074a9305c04054cdebb9e9aad2d818052b07091de1f20cad0bbac34ffb52L85-R143): Refactors reproducible build targets, centralizes environment variables for reproducibility, and updates Docker build arguments for `x86_64` and `aarch64` architectures.
* [`Dockerfile.reproducible`](diffhunk://#diff-587298ff141278ce3be7c54a559f9f31472cc5b384e285e2105b3dee319ba31dL1-R24): Updates the base Rust image to version 1.86, removes hardcoded reproducibility settings, and delegates build logic to the `Makefile`.
* Switch to using jemalloc-sys from Debian repos instead of building it from source. A Debian version is [reproducible](https://tests.reproducible-builds.org/debian/rb-pkg/trixie/amd64/jemalloc.html) which is [hard to achieve](https://github.com/NixOS/nixpkgs/issues/380852) if you build it from source.
### Profile Removal:
* [`Cargo.toml`](diffhunk://#diff-2e9d962a08321605940b5a657135052fbcef87b5e360662bb527c96d9a615542L289-L295): Removes the `reproducible` profile, simplifying build configurations and relying on external tooling for reproducibility.
Co-Authored-By: Moe Mahhouk <mohammed-mahhouk@hotmail.com>
Co-Authored-By: chonghe <44791194+chong-he@users.noreply.github.com>
Co-Authored-By: Michael Sproul <michaelsproul@users.noreply.github.com>
Use the recently published `context_deserialize` and remove it from Lighthouse
Co-Authored-By: Mac L <mjladson@pm.me>
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
Fixes#7785
- [x] Update all integration tests with >1 files to follow the `main` pattern.
- [x] `crypto/eth2_key_derivation/tests`
- [x] `crypto/eth2_keystore/tests`
- [x] `crypto/eth2_wallet/tests`
- [x] `slasher/tests`
- [x] `common/eth2_interop_keypairs/tests`
- [x] `beacon_node/lighthouse_network/tests`
- [x] Set `debug_assertions` to false on `.vscode/settings.json`.
- [x] Document how to make rust analyzer work on integration tests files. In `book/src/contributing_setup.md`
---
Tracking a `rust-analyzer.toml` with settings like the one provided in `.vscode/settings.json` would be nicer. But this is not possible yet. For now, that config should be a good enough indicator for devs using editors different to VSCode.
Co-Authored-By: Daniel Ramirez-Chiquillo <hi@danielrachi.com>
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
Consolidate our property-testing around `proptest`. This PR was written with Copilot and manually tweaked.
Co-Authored-By: Michael Sproul <michael@sproul.xyz>
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
When developing locally with kurtosis a typical dev workflow is:
loop:
- Build local lighthouse docker image
- Run kurtosis
- Observe bug
- Fix code
The docker build step would download and build all crates. Docker docs suggests an optimization to cache build artifacts, see https://docs.docker.com/build/cache/optimize/#use-cache-mounts
I have tested and it's like building Lighthouse outside of a docker environment 🤤 The docker build time after changing one line in the top beacon_node crate is 50 seconds on my local machine ❤️
The release path is un-affected. Do you have worries this can affect the output of the release binaries? This is too good of an improvement to keep it in a separate Dockerfile.
Co-Authored-By: dapplion <35266934+dapplion@users.noreply.github.com>
Update `reqwest` to 0.12 so we only depend on a single version. This should slightly improve compile times and reduce binary bloat.
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
Currently whenever we build the `Dockerfile` file for local development using kurtosis , it recompiles everything on my laptop, even if no changes are made. This takes about 120 seconds on my laptop (might be faster on others).
Conservatively, I created a new Dockerfile.dev, so that the original file is kept the same, even though its pretty similar.
This uses `--mount-type=cache` saving the target and registry folder across builds.
**Usage**
```sh
docker build -f Dockerfile.dev -t lighthouse:dev .
```
Co-Authored-By: Kevaundray Wedderburn <kevtheappdev@gmail.com>
This is a `tracing`-driven optimisation. While investigating why Lighthouse is slow to send `newPayload`, I found a suspicious 13ms of computation on the hot path in `gossip_block_into_execution_pending_block_slashable`:
<img width="1998" height="1022" alt="headercalc" src="https://github.com/user-attachments/assets/e4f88c1a-da23-47b4-b533-cf5479a1c55c" />
Looking at the current implementation we can see that the _only_ thing that happens prior to calling into `from_gossip_verified_block` is the calculation of a `header`. We first call `SignatureVerifiedBlock::from_gossip_verified_block_check_slashable`:
261322c3e3/beacon_node/beacon_chain/src/block_verification.rs (L1075-L1076)
Which is where the `header` is calculated prior to calling `from_gossip_verified_block`:
261322c3e3/beacon_node/beacon_chain/src/block_verification.rs (L1224-L1226)
Notice that the `header` is _only_ used in the case of an error, yet we spend time computing it every time!
This PR moves the calculation of the header (which involves hashing the whole beacon block, including the execution payload), into the error case. We take a cheap clone of the `Arc`'d beacon block on the hot path, and use this for calculating the header _only_ in the case an error actually occurs. This shaves 10-20ms off our pre-newPayload delays, and 10-20ms off every block processing 🎉
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
We want to not require checkpoint sync starts to include the required custody data columns, and instead fetch them from p2p.
Closes https://github.com/sigp/lighthouse/issues/6837
The checkpoint sync slot can:
1. Be the first slot in the epoch, such that the epoch of the block == the start checkpoint epoch
2. Be in an epoch prior to the start checkpoint epoch
In both cases backfill sync already fetches that epoch worth of blocks with current code. This PR modifies the backfill import filter function to allow to re-importing the oldest block slot in the DB.
I feel this solution is sufficient unless I'm missing something. ~~I have not tested this yet!~~ Michael has tested this and it works.
Co-Authored-By: dapplion <35266934+dapplion@users.noreply.github.com>
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
N/A
The difference is computed by taking the difference of expected with received. We were doing the inverse.
Thanks to Yassine for finding the issue.
Co-Authored-By: Pawan Dhananjay <pawandhananjay@gmail.com>
Take 2 of #8390.
Fixes the race condition properly instead of propagating the error. I think this is a better alternative, and doesn't seem to look that bad.
* Lift node id loading or generation from `NetworkService ` startup to the `ClientBuilder`, so that it can be used to compute custody columns for the beacon chain without waiting for Network bootstrap.
I've considered and implemented a few alternatives:
1. passing `node_id` to beacon chain builder and compute columns when creating `CustodyContext`. This approach isn't good for separation of concerns and isn't great for testability
2. passing `ordered_custody_groups` to beacon chain. `CustodyContext` only uses this to compute ordered custody columns, so we might as well lift this logic out, so we don't have to do error handling in `CustodyContext` construction. Less tests to update;.
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
Addressed this comment here: https://github.com/sigp/lighthouse/issues/6837#issuecomment-3509209465
Lighthouse can only checkpoint sync from a server that can serve blob sidecars, which means they need to be at least custdoying 50% of columns (semi-supernodes)
This PR lifts this constraint, as blob sidecar endpoint is getting deprecated in Fulu, and we plan to fetch the checkpoint data columns from peers (#6837)
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
Fix the span on execution payload verification (newPayload), by creating a new span rather than using the parent span. Using the parent span was incorrectly associating the time spent verifying the payload with `from_signature_verified_components`.
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
Fix an issue detected by @jimmygchen that occurs when checkpoint sync is aborted midway and then later restarted.
The characteristic error is something like:
> Nov 13 00:51:35.832 ERROR Database write failed error: Hdiff(LessThanStart(Slot(1728288), Slot(1728320))), action: "reverting blob DB changes"
Nov 13 00:51:35.833 WARN Hot DB pruning failed error: DBError(HotColdDBError(Rollback))
This issue has existed since v7.1.0.
Delete snapshot/diff in the case where `hot_storage_strategy` fails.
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
#6022
Migrate the `execution_engine_integration` tests to the `alloy` ecosystem. This removes the last remaining `ethers` dependencies
Co-Authored-By: Mac L <mjladson@pm.me>
Part of a fork-choice tech debt clean-up https://github.com/sigp/lighthouse/issues/8325https://github.com/sigp/lighthouse/issues/7089 (non-finalized checkpoint sync) changes the meaning of the checkpoints inside fork-choice. It turns out that we persist the justified and finalized checkpoints **twice** in fork-choice
1. Inside the fork-choice store
2. Inside the proto-array
There's no reason for 2. except for making the function signature of some methods smallers. It's not consistent with the rest of the crate, because in some functions we pass the external variable of time (current_slot) via args, but then read the finalized checkpoint from the internal state. Passing both variables as args makes fork-choice easier to reason about at the cost of a few extra lines.
Remove the unnecessary state (`justified_checkpoint`, `finalized_checkpoint`) inside `ProtoArray`, to make it easier to reason about.
Co-Authored-By: dapplion <35266934+dapplion@users.noreply.github.com>
Co-Authored-By: Michael Sproul <michaelsproul@users.noreply.github.com>
State advances were observed as especially slow on pre-Fulu networks (mainnet).
The reason being: we were doing an extra epoch of state advance because of code that should only have been running after Fulu, when proposer shufflings are determined with lookahead.
Only attempt to cache the _next epoch_ shuffling if the state's slot determines it (this will only be true post-Fulu). Reusing the logic for `proposer_shuffling_decision_slot` avoids having to repeat the fiddly logic about the Fulu fork epoch itself.
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
since block and blob both start with `bl`, it was not clear how to differentiate between `blbroots_queue` and `bbroots_queue`
After renaming, there also seems to be a discrepancy
Co-Authored-By: Kevaundray Wedderburn <kevtheappdev@gmail.com>
Debugging https://github.com/sigp/lighthouse/issues/8104 it would have been helpful to quickly see in the logs that a specific block was submitted into the HTTP API.
Because we want to optimize the block root computation we don't include it in the logs, and just log the block slot. I believe we can take a minute performance hit to have the block root in all the logs during block publishing.
Co-Authored-By: dapplion <35266934+dapplion@users.noreply.github.com>
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
#6022
Use `alloy_rpc_types::Transaction` to replace the `ethers_core::Transaction` inside the execution block generator.
Co-Authored-By: Mac L <mjladson@pm.me>
FIx flaky tests that depends on timing. Previously the test processes all 128 columns and expect reconstruction to happen after all columns are processed. There is a race here, and reconstruction could be triggered before all columns are processed.
I've updated the tests to process 64 columns, just enough for reconstruction and wait for 50ms for reconstruction to be triggered.
This PR requires the change made in https://github.com/sigp/lighthouse/pull/8194 for the test to pass consistently (blob count set to 1 for all blocks instead of random blob count between 0..max)
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
Co-Authored-By: Jimmy Chen <jimmy@sigmaprime.io>
Another good candidate for publishing separately from Lighthouse is `sensitive_url` as it's a general utility crate and not related to Ethereum. This PR prepares it to be spun out into its own crate.
I've made the `full` field on `SensitiveUrl` private and instead provided an explicit getter called `.expose_full()`. It's a bit ugly for the diff but I prefer the explicit nature of the getter.
I've also added some extra tests and doc strings along with feature gating `Serialize` and `Deserialize` implementations behind the `serde` feature.
Co-Authored-By: Mac L <mjladson@pm.me>
### Downgrade a non error to `Debug`
I noticed this error on one of our hoodi nodes:
```
Nov 04 05:13:38.892 ERROR Error during data column reconstruction block_root: 0x4271b9efae7deccec3989bd2418e998b83ce8144210c2b17200abb62b7951190, error: DuplicateFullyImported(0x4271b9efae7deccec3989bd2418e998b83ce8144210c2b17200abb62b7951190)
```
This shouldn't be logged as an error and it's due to a normal race condition, and it doesn't impact the node negatively.
### Remove spammy logs
This logs is filling up the log files quite quickly and it is also something we'd expect during normal operation - getting columns via EL before gossip. We haven't found this debug log to be useful, so I propose we remove it to avoid spamming debug logs.
```
Received already available column sidecar. Ignoring the column sidecar
```
In the process of removing this, I noticed we aren't propagating the validation result, which I think we should so I've added this. The impact should be quite minimal - the message will stay in the gossip memcache for a bit longer but should be evicted in the next heartbeat.
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
while working on this #7892 @michaelsproul pointed it might be a good metric to measure the delay from start of the slot instead of the current `slot_duration / 3`, since the attestations duties start before the `1/3rd` mark now with the change in the link PR.
Co-Authored-By: hopinheimer <knmanas6@gmail.com>
Co-Authored-By: hopinheimer <48147533+hopinheimer@users.noreply.github.com>
This is an optimisation targeted at Fulu networks in non-finality.
While debugging on Holesky, we found that `state_root_at_slot` was being called from `prepare_beacon_proposer` a lot, for the finalized state:
2c9b670f5d/beacon_node/http_api/src/lib.rs (L3860-L3861)
This was causing `prepare_beacon_proposer` calls to take upwards of 5 seconds, sometimes 10 seconds, because it would trigger _multiple_ beacon state loads in order to iterate back to the finalized slot. Ideally, loading the finalized state should be quick because we keep it cached in the state cache (technically we keep the split state, but they usually coincide). Instead we are computing the finalized state root separately (slow), and then loading the state from the cache (fast).
Although it would be possible to make the API faster by removing the `state_root_at_slot` call, I believe it's simpler to change `state_root_at_slot` itself and remove the footgun. Devs rightly expect operations involving the finalized state to be fast.
Co-Authored-By: Michael Sproul <michael@sigmaprime.io>
`beacon-chain-tests` is now regularly taking 1h+ on CI since Fulu fork was added.
This PR attemtpts to reduce the test time by bringing down the number of blobs generated in tests - instead of generating 0..max_blobs, the generator now generates 0..1 blobs by default, and this can be modified by setting `harness.execution_block_generator.set_min_blob_count(n)`.
Note: The blobs are pre-generated and doesn't require too much CPU to generate however processing a larger number of them on the beacon chain does take a lot of time.
This PR also include a few other small improvements
- Our slowest test (`chain_segment_varying_chunk_size`) runs 3x faster in Fulu just by reusing chain segments
- Avoid re-running fork specific tests on all forks
- Fix a bunch of tests that depends on the harness's existing random blob generation, which is fragile
beacon chain test time on test machine is **~2x** faster:
### `unstable`
```
Summary [ 751.586s] 291 tests run: 291 passed (13 slow), 0 skipped
```
### this branch
```
Summary [ 373.792s] 291 tests run: 291 passed (2 slow), 0 skipped
```
The next set of tests to optimise is the ones that use [`get_chain_segment`](77a9af96de/beacon_node/beacon_chain/tests/block_verification.rs (L45)), as it by default build 320 blocks with supernode - an easy optimisation would be to build these blocks with cgc = 8 for tests that only require fullnodes.
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
Co-Authored-By: Jimmy Chen <jimmy@sigmaprime.io>
N/A
Includes the following unmerged PRs:
- #8344
- #8335
- #8339
This PR should be merged after all above PRs are merged.
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
Co-Authored-By: Jimmy Chen <jimmy@sigmaprime.io>
Custody backfill sync has a bug when we request columns from more than one peer per batch. The fix here ensures we wait for all requests to be completed before performing verification and importing the responses.
I've also added an endpoint `lighthouse/custody/backfill` that resets a nodes earliest available data column to the current epoch so that custody backfill can be triggered. This endpoint is needed to rescue any nodes that may have missing columns due to the custody backfill sync bug without requiring a full re-sync.
Co-Authored-By: Eitan Seri- Levi <eserilev@gmail.com>
Co-Authored-By: Eitan Seri-Levi <eserilev@ucsc.edu>
Co-Authored-By: Jimmy Chen <jchen.tc@gmail.com>
Co-Authored-By: Michael Sproul <michaelsproul@users.noreply.github.com>
