Fix out-of-order sample loss by making remote write synchronous#109
Open
cnewkirk wants to merge 1 commit intoOpenNMS:masterfrom
Open
Fix out-of-order sample loss by making remote write synchronous#109cnewkirk wants to merge 1 commit intoOpenNMS:masterfrom
cnewkirk wants to merge 1 commit intoOpenNMS:masterfrom
Conversation
The store() method previously fired HTTP writes asynchronously via executeAsync() and returned immediately. When the RingBuffer's multiple worker threads dispatched consecutive batches containing samples for the same series, the async HTTP requests could arrive at the remote write endpoint out of timestamp order, causing the backend to reject the stale samples as out-of-order. This change makes store() block until the HTTP write completes, ensuring the ring buffer worker thread does not process the next batch until the current write has landed. This preserves per-series timestamp ordering across consecutive WriteRequests as required by the Prometheus Remote Write spec. Additionally fixes a bug where samplesLost incorrectly counted unfiltered samples (including NaN) instead of the actual samples that were attempted. Validated via A/B E2E testing against Thanos Receive: - Baseline (async): 14 out-of-order / 5,262 appended (0.27% loss) - Fix (sync): 0 out-of-order / 5,264 appended (0.00% loss) - Throughput: identical (~5,260 samples over equal soak periods) - All 45 smoke tests passing on both runs Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
|
Created https://opennms.atlassian.net/browse/NMS-19647 for this one. |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The store() method previously fired HTTP writes asynchronously via executeAsync() and returned immediately. When the RingBuffer's multiple worker threads dispatched consecutive batches containing samples for the same series, the async HTTP requests could arrive at the remote write endpoint out of timestamp order, causing the backend to reject the stale samples as out-of-order.
This change makes store() block until the HTTP write completes, ensuring the ring buffer worker thread does not process the next batch until the current write has landed. This preserves per-series timestamp ordering across consecutive WriteRequests as required by the Prometheus Remote Write spec.
Additionally fixes a bug where samplesLost incorrectly counted unfiltered samples (including NaN) instead of the actual samples that were attempted.
Validated via A/B E2E testing against Thanos Receive: