BewareMyPower commented on issue #24697: URL: https://github.com/apache/pulsar/issues/24697#issuecomment-3292474234
> After changing managedLedgerNewEntriesCheckDelayInMillis=0, this issue not reproduced yet. This is an interesting point, the main difference is that when there is no entries available, when and where to call `checkForNewEntries` again: - 0: call it immediately in the current thread (usually the `broker-topics-xxx` worker thread of `PersistentTopic`). Mostly it should just keep the `OpReadEntry` in `waitingReadOp` and wait for new messages were written. - default (10): schedule the method in managed ledger's executor. Though it's still hard to say it's the direct cause, it's reasonable to assume that the `add OpReadEntry to waitingReadOp -> notified by notifyEntriesAvailable` path is more safe -- 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]
