Today we have seen few of the Spark jobs are running/hanging. The jobs are already finished, however the spark AMs were not killed by the RM. Could you please help me in checking the issue.
I have checked the AM logs. The last time the log was updated 10 hours back.
I have checked the Yarn RM logs. Here also the last time the log was updated 10 hours back. Since then, there is no progress in the logs. However, the containers/jobs are in RUNNING state in Yarn UI holding some containers impacting other jobs to run
I don't see any alerts in Ambari. The rest of the jobs are running good. I have killed these jobs and re ran. The jobs went successful now. Don't understand how the jobs were stuck.