Repository navigation
Conversation
Read cost grows with retained data, so a read gate measured on a nearly empty server is optimistic. --prefill runs an ingest-only mixed phase after warmup, waits for the ingest queue and rollup backlog to drain, and records a prefill-settled gate. Prefill rows are excluded from totals, peaks and topology selection.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Why
Read cost grows with retained data. Two Hetzner ccx33 runs at the same 48k items/s mixed ingest with 20 UI reads/s on
v2026.11.9-alphaandv2026.11.8-alphagave very different results:In the prefilled run, every failure was a 10-second timeout on
/api/queries/exemplars. A read gate measured on a nearly empty server therefore overstates interactive performance.What
--prefill <duration>and--prefill-processes <n>(default 16) add an ingest-only mixed phase after warmup. Prefill runs without read load and uses the preset's workers and rate.pitot_ingest_queue_depthandpitot_rollup_backlog_chunksare exported and read zero in two consecutive scrapes. A missing gauge never counts as drained. The wait is skipped for compaction backlog, which counts underfilled groups and need not reach zero.prefill-settledquality gate fails when the backlog does not drain.Tests
go test ./...passes with 177 tests. New tests cover the plan, the gate, the settle loop (two clean scrapes, a missing gauge, cancellation), flag plumbing and the workload hash.