lweitzendorf opened a new pull request, #3204: URL: https://github.com/apache/jackrabbit-oak/pull/3204
https://issues.apache.org/jira/browse/OAK-12449 ## Summary When the dispatch queue already holds `DISPATCH_QUEUE_MAX_SIZE` tasks, `SegmentPreloader`'s dispatch executor drops new tasks in `execute()`. Those tasks were already registered in `inProgressPrefetch`, but since they never run, `afterExecute()` never removes them. The map grows under sustained preload pressure, and the affected segments are never preloaded again because they appear to be in progress. The fix calls `clearInProgressTask(command)` when a task is dropped. ## Tests `SegmentPreloaderTest#droppedDispatchTaskIsNotLeftInProgress` blocks the dispatcher, fills the queue to its limit, submits one more task and checks that `inProgressPrefetch` is empty. It fails without the fix. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
