Repository navigation
Conversation
Drill-down ranks every retained trace in its window, so it times out once the server holds data. Propose exactly maintained per-bucket rank lists in the read cache, and add the filebench scale benchmark that measures the current cost.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Drill-down ranks every retained trace in its window to return the top 21, so its cost grows with retained data. A server that has ingested for a while times out on
/api/queries/exemplars. This draft proposes exactly maintained per-bucket rank lists in the read cache, so drill-down cost depends on window length, not retained traces.Evidence
v2026.11.9-alpha, 48k items/s mixed ingest with 20 UI reads/s: a fresh server passes at 20/20 ops/s. After a 10-minute prefill (about 180M rows) it completes 2.6/20 ops/s. All 1,716 failures are exemplar 504s.filebenchscale benchmark, fully cached: 400k traces take 21 ms (single-service) and 93 ms (two-service); 1.6M traces take 197 ms and 633 ms.Contents
docs/proposals/0012-bounded-trace-ranking.md: design, exactness argument, maintenance under publication, compaction and retention, four phases with gates (the last is a full-day prefill on Hetzner), alternatives and risks.internal/engine/trace_candidate_scale_bench_test.go: build tagfilebench.PITOT_SCALE_BATCHESsets cached batches of 20k traces;PITOT_SCALE_MIXED=1spreads traces over two services.No product code changes. Related: #480, labstack/pitot-bench#14.