gnosis/lighthouse
1
0
Fork 0
mirror of https://github.com/sigp/lighthouse.git synced 2026-03-19 21:04:41 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix the getBlobs metric and ensure it is recorded promptly to prevent miscounts (#7188)

Browse Source 
From testing conducted by Sunnyside Labs, they noticed that the "expected blobs" are quite low on bandwidth constrained nodes. This observation revealed that we don't record the `beacon_blobs_from_el_expected_total` metric at all if the EL doesn't return any response. The fetch blobs function returns without recording the metric.

To fix this, I've moved `BLOBS_FROM_EL_EXPECTED_TOTAL` and `BLOBS_FROM_EL_RECEIVED_TOTAL` to as early as possible, to make the metric more accurate.
This commit is contained in:
Jimmy Chen
2025-04-04 20:01:39 +11:00
committed by GitHub
parent 57abffcd99
commit 6a75f24ab1
2 changed files with 24 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
beacon_node/beacon_chain/src/fetch_blobs.rs
View File

@@ -13,7 +13,7 @@ use crate::observed_data_sidecars::DoNotObserve;
use crate::{metrics, AvailabilityProcessingStatus, BeaconChain, BeaconChainTypes, BlockError};
use execution_layer::json_structures::BlobAndProofV1;
use execution_layer::Error as ExecutionLayerError;
use metrics::{inc_counter, inc_counter_by, TryExt};
use metrics::{inc_counter, TryExt};
use ssz_types::FixedVector;
use state_processing::per_block_processing::deneb::kzg_commitment_to_versioned_hash;
use std::sync::Arc;
@@ -73,13 +73,20 @@ pub async fn fetch_and_process_engine_blobs<T: BeaconChainTypes>(
.as_ref()
.ok_or(FetchEngineBlobError::ExecutionLayerMissing)?;
metrics::observe(&metrics::BLOBS_FROM_EL_EXPECTED, num_expected_blobs as f64);
debug!(num_expected_blobs, "Fetching blobs from the EL");
let response = execution_layer
.get_blobs(versioned_hashes)
.await
.inspect_err(|_| {
inc_counter(&metrics::BLOBS_FROM_EL_ERROR_TOTAL);
})
.map_err(FetchEngineBlobError::RequestFailed)?;
if response.is_empty() || response.iter().all(|opt| opt.is_none()) {
let num_fetched_blobs = response.iter().filter(|opt| opt.is_some()).count();
metrics::observe(&metrics::BLOBS_FROM_EL_RECEIVED, num_fetched_blobs as f64);
if num_fetched_blobs == 0 {
debug!(num_expected_blobs, "No blobs fetched from the EL");
inc_counter(&metrics::BLOBS_FROM_EL_MISS_TOTAL);
return Ok(None);
@@ -99,20 +106,6 @@ pub async fn fetch_and_process_engine_blobs<T: BeaconChainTypes>(
&chain.spec,
)?;
let num_fetched_blobs = fixed_blob_sidecar_list
.iter()
.filter(|b| b.is_some())
.count();
inc_counter_by(
&metrics::BLOBS_FROM_EL_EXPECTED_TOTAL,
num_expected_blobs as u64,
);
inc_counter_by(
&metrics::BLOBS_FROM_EL_RECEIVED_TOTAL,
num_fetched_blobs as u64,
);
// Gossip verify blobs before publishing. This prevents blobs with invalid KZG proofs from
// the EL making it into the data availability checker. We do not immediately add these
// blobs to the observed blobs/columns cache because we want to allow blobs/columns to arrive on gossip