Wang1rrr opened a new issue, #6977: URL: https://github.com/apache/rocketmq-dashboard/issues/6977
### Description `InstanceResourceCountRunner.close()` calls `shutdownNow()` but ignores the queued tasks it returns. Those tasks are `FutureTask`s removed from the executor queue without being completed or cancelled. A `countAll` request already waiting on them therefore remains blocked until its full batch timeout even though the runner has closed and the tasks can never execute. ### Reproduction Use a runner with one worker, queue one additional count, and block the running count on an interruptible latch. Close the runner while both counts are outstanding. The active task is interrupted, but the request continues waiting on the drained queued future until its configured deadline. ### Expected behavior Cancel tasks drained by shutdown so waiting requests promptly observe interruption. No queued provider call should run after close. ### Version `rocketmq-studio` at c34535db0819d02f68e50a856cb3d2614f0dc5dd. I plan a focused shutdown fix with a latch-controlled regression. -- 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]
