zhijiangW commented on pull request #11687:
URL: https://github.com/apache/flink/pull/11687#issuecomment-626507028
@pnowojski I already addressed the last nit comments. Besides that I also
adjusted another two places.
1. In `StreamTask#beforeInvoke`, I take the following way instead to
guarantee the partition request execute by task thread.
```
CompletableFuture.allOf(futures).thenRun(() ->
mailboxProcessor.getMainMailboxExecutor().execute(
() -> {
for (InputGate inputGate :
inputGates) {
inputGate.requestPartitions();
}
}, "Input gates request partitions"));
```
2. In `RecoveredInputChannel#getNextRecoveredStateBuffer`, I move the future
complete and related processes out of the `synchronized (receivedBuffers)` to
avoid potential deadlock issue.
I will merge the PR after azure pass if you have no further concerns.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
