AM restart in a other node makes SparkSQL jobs into a state of feign death

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

AM restart in a other node makes SparkSQL jobs into a state of feign death

Bang Xiao
I run "spark-sql  --master yarn --deploy-mode client -f 'SQLs' " in shell,
The application  is stuck when the AM is down and restart in other nodes. It
seems the driver wait for the next sql. Is this a bug?In my opinion,Either
the application execute the failed sql or exit with a failure when the AM
restart。



--
Sent from: http://apache-spark-user-list.1001560.n3.nabble.com/

---------------------------------------------------------------------
To unsubscribe e-mail: [hidden email]