Hello guys, we have done a upgrade from flink 13 to flink 16 now, we see a significant increase for our flink taskmanager backpressure time, is there some conf in flink 16 to do tunning so we can optimize le backpressure time ?🙏 here is the metric we used to observe backpressure

flink_taskmanager_job_task_backPressuredTimeMsPerSecond

Have you identified which operator is causing the back pressure? You can look on the Flink dashboard

We have identified the operator, it's a operator which is using event time window with 30 minutes gap configured, but it's the same operator we have used it before

I am not sure there is a general config to fix this for 1.16. You will need to root cause it first. If the operator is not backpressured by external system (sink), then recommend looking at the FlameGraph to see what CPU is doing

ok, thanks, we'll do it