Skip to content

perf: remove parser-stage guard to expand stats_query_range fast path #352

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
szibis merged 9 commits into from
May 13, 2026
Merged

perf: remove parser-stage guard to expand stats_query_range fast path #352

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
305557a
docs: add metric fast-path expansion design spec
szibis May 12, 2026
9b6fc79
docs: add metric fast-path implementation plan
szibis May 12, 2026
8bd7f81
test: add failing unit tests for parser-stage fast-path gate removal
szibis May 12, 2026
e67ae63
perf: remove parser-stage guard from shouldUseManualRangeMetricCompat
szibis May 12, 2026
149cbe5
fix: correct VL syntax in shouldUseManualRangeMetricCompat comment
szibis May 12, 2026
a4cd0c5
test: add integration tests for parser-stage rate tumbling/sliding ro…
szibis May 12, 2026
4635e96
test: align stats_query_range stub with actual VL response format
szibis May 12, 2026
7ec5e78
test: update sliding-window test for PR #350 collectRangeMetricHits f…
szibis May 12, 2026
af0dd93
docs(changelog): add entry for parser-stage guard removal
szibis May 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0

### Performance

- perf(range_metric): remove parser-stage guard from `shouldUseManualRangeMetricCompat` — `rate`, `bytes_rate`, `count_over_time`, and `bytes_over_time` queries containing `| json` or `| logfmt` parser stages now use VL native `stats_query_range` for tumbling windows (`range == step`), previously forced onto the raw log-fetch slow path regardless of window configuration.
- perf(range_metric): add `stats_query_range` fast path for `sum by (...) (count_over_time/rate({...}[W]))` — replaces raw NDJSON log scan (39% CPU cold proxy) with VL pre-aggregated Prometheus buckets via `| stats by (...) count() as c`. Throughput improvement: 44→126 req/s (c=10) and 33→139 req/s (c=100) on heavy workload.
- perf(range_metric): extend fast path to `bytes_over_time`/`bytes_rate` using `| stats by (...) sum_len(_msg) as c` — eliminates 4664ms P50 bottleneck for byte-rate metric queries at cold concurrency.
- perf(patterns): pool `strings.Builder` in `patternLineTokenizer.Join` via `patternJoinBuilderPool` — eliminates per-log-entry heap allocation that was 6237 MB flat (15.56%) and driving `bytes.growSlice` cascade at cold concurrency.
Expand Down
Loading