mirror of
https://github.com/rustfs/rustfs.git
synced 2026-08-21 20:06:37 +00:00
perf(filemeta): phase-1~3 rename_data metadata optimization (#3011)
* chore(perf): harden amd64 profiling benchmark flow * fix(profiling): isolate bench buckets and map protobuf conflict * perf: avoid blocking owned local writes * style: format profile admin handler * docs: clarify observability trace validation * perf: reduce mkdir overhead on local writes * perf: add rename_data meta microbenchmark * perf(filemeta): fast-path data_dir decode in version meta * perf(filemeta): collapse data-dir lookup into one scan * perf(filemeta): reduce scan allocs and refresh meta bench * perf(ecstore): skip mkdir path on read-only open * perf(filemeta): single-pass unshared data-dir scan * perf(filemeta): add two-key inline remove fast path * perf(filemeta): compare remove-two keys by bytes first * bench(ecstore): add remove_two-only micro benchmark * bench(ecstore): stabilize rename_data meta benchmark timing * bench(ecstore): align rename_data path with remove_two * perf(filemeta): avoid uuid string alloc in remove_two * perf(filemeta): add fast-path for empty inline data * perf(filemeta): streamline add_version match branch * perf(filemeta): fast-return remove_key on miss * perf(filemeta): speed up add_version insertion lookup * style(ecstore): normalize formatting in perf-tuning files * refactor(filemeta): unify inline data removal paths
This commit is contained in:
@@ -54,6 +54,46 @@ Strict mode is available when you explicitly want `/health/ready == 200` as the
|
||||
WAIT_PROBE_MODE=ready ./scripts/run_four_node_cluster_failover_bench.sh
|
||||
```
|
||||
|
||||
### Profiling + Trace Validation
|
||||
|
||||
The profiling-focused 4-node compose keeps profiling enabled and points RustFS
|
||||
to an OTLP/HTTP collector endpoint:
|
||||
|
||||
```bash
|
||||
docker compose -f .docker/compose/docker-compose.cluster.local-build.profiling-amd64.yml up -d
|
||||
```
|
||||
|
||||
Important behavior notes:
|
||||
|
||||
- `RUSTFS_OBS_ENDPOINT` is the OTLP/HTTP base URL. RustFS automatically sends
|
||||
traces to `/v1/traces`, metrics to `/v1/metrics`, and logs to `/v1/logs`.
|
||||
- Startup usually produces logs and metrics first. That does not guarantee
|
||||
visible traces yet.
|
||||
- Trace data becomes obvious only after real HTTP/S3/gRPC requests hit RustFS.
|
||||
- `RUSTFS_OBS_LOGGER_LEVEL=info` keeps the top-level request span but filters
|
||||
many nested `debug` spans. If Tempo/Jaeger looks sparse, retry with
|
||||
`RUSTFS_OBS_LOGGER_LEVEL=debug` before suspecting the collector.
|
||||
|
||||
Minimal trace verification flow:
|
||||
|
||||
```bash
|
||||
# 1. Start the profiling compose with richer span visibility.
|
||||
RUSTFS_OBS_LOGGER_LEVEL=debug \
|
||||
docker compose -f .docker/compose/docker-compose.cluster.local-build.profiling-amd64.yml up -d
|
||||
|
||||
# 2. Generate real request traffic after startup.
|
||||
curl -I http://127.0.0.1:9000/health
|
||||
curl -I http://127.0.0.1:9000/health/ready
|
||||
|
||||
# 3. Then inspect Tempo or Jaeger.
|
||||
# Grafana: http://localhost:3000
|
||||
# Jaeger: http://localhost:16686
|
||||
```
|
||||
|
||||
If logs and metrics are present but traces are sparse, the most common cause is
|
||||
"no real request traffic yet" or "`info` level filtered nested spans", not an
|
||||
OTLP routing failure.
|
||||
|
||||
### (Deprecated) Minimal Observability
|
||||
|
||||
```bash
|
||||
|
||||
Reference in New Issue
Block a user