Hi Yun Tang，
I restart the job ，it has been running for 12 hours so far. There has
been no failure to execute checkpoint so far, but some checkpoint execution
takes a long time.
<http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/file/t1922/checkpoint_time.png>
This is the detailed information for performing checkpoint.Time is mainly
spent in the synchronization phase of executing checkpoint.The detailed
jstack log of this subtask is as follows：
<http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/file/t1922/jstack_log.png>
I guess the checkpoint synchronization phase takes too long to execute, is
it because I set the .cleanupInRocksdbCompactFilter(1000) parameter value
too large.
Best
Yang
--
Sent from: http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/
