Repository navigation
feat: configure WebSocket worker delivery timeout - #106
Merged
Merged
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A throttled worker can legitimately wait longer than the cluster's hardcoded 30-second WebSocket acknowledgement deadline. In ONE MQTT catch-up, the rolling limiter can wait up to 60 seconds even though the source reservation is configured for 90 seconds; changing the source timeout does not change the worker tracker.
Add optional
workerDeliveryTimeoutMstoFlowcoreDataPumpCluster, preserving the 30-second default. Validate integer timer bounds before creating cluster state. The option affects only the WebSocket acknowledgement tracker; NATS, source reservation behavior and retry/late-ACK semantics stay unchanged. Document processing budget, source reservation margin, and the fact that expiry cannot cancel worker side effects.Validation: baseline implementation fails 7 of the new regression cases; the corrected focused suite passes 10 tests. Full suite: 166 tests / 569 assertions passed. Lint, typecheck, formatting and JS/declaration build passed with Bun 1.4.2 locally; hosted CI pins 1.3.14. Independent spec/safety review found no blockers. Integration infrastructure tests remain for hosted CI.
This library PR alone changes no production configuration. Release/consumer wiring and synthetic backlog qualification are required before ONE main cutover.