Instrument tracing spans for block processing and import (#7816)

#7815

- removes all existing spans, so some span fields that appear in logs like `service_name` may be lost.
- instruments a few key code paths in the beacon node, starting from **root spans** named below:

* Gossip block and blobs
* `process_gossip_data_column_sidecar`
* `process_gossip_blob`
* `process_gossip_block`
* Rpc block and blobs
* `process_rpc_block`
* `process_rpc_blobs`
* `process_rpc_custody_columns`
* Rpc blocks (range and backfill)
* `process_chain_segment`
* `PendingComponents` lifecycle
* `pending_components`

To test locally:
* Run Grafana and Tempo with https://github.com/sigp/lighthouse-metrics/pull/57
* Run Lighthouse BN with `--telemetry-collector-url http://localhost:4317`

Some captured traces can be found here: https://hackmd.io/@jimmygchen/r1sLOxPPeg

Removing the old spans seem to have reduced the memory usage quite a lot - i think we were using them on long running tasks and too excessively:
<img width="910" height="495" alt="image" src="https://github.com/user-attachments/assets/5208bbe4-53b2-4ead-bc71-0b782c788669" />
This commit is contained in:
Jimmy Chen
2025-08-08 15:32:22 +10:00
committed by GitHub
parent 6dfab22267
commit 40c2fd5ff4
52 changed files with 633 additions and 1164 deletions

View File

@@ -13,7 +13,7 @@ use std::cmp::Ordering;
use std::collections::BTreeSet;
use std::marker::PhantomData;
use std::time::Duration;
use tracing::{debug, warn};
use tracing::{debug, instrument, warn};
use types::{
consts::bellatrix::INTERVALS_PER_SLOT, AbstractExecPayload, AttestationShufflingId,
AttesterSlashingRef, BeaconBlockRef, BeaconState, BeaconStateError, ChainSpec, Checkpoint,
@@ -472,6 +472,7 @@ where
/// Is equivalent to:
///
/// https://github.com/ethereum/eth2.0-specs/blob/v0.12.1/specs/phase0/fork-choice.md#get_head
#[instrument(skip_all, level = "debug")]
pub fn get_head(
&mut self,
system_time_current_slot: Slot,
@@ -646,6 +647,12 @@ where
/// The supplied block **must** pass the `state_transition` function as it will not be run
/// here.
#[allow(clippy::too_many_arguments)]
#[instrument(
name = "fork_choice_on_block",
skip_all,
fields(
fork_choice_block_delay = ?block_delay
))]
pub fn on_block<Payload: AbstractExecPayload<E>>(
&mut self,
system_time_current_slot: Slot,