KaiXinXIaoLei created SPARK-11298:
-------------------------------------
Summary: When driver sends message "GetExecutorLossReason" to AM,
the AM stops.
Key: SPARK-11298
URL: https://issues.apache.org/jira/browse/SPARK-11298
Project: Spark
Issue Type: Bug
Components: Spark Core
Affects Versions: 1.6.0
Reporter: KaiXinXIaoLei
Fix For: 1.6.0, 2+
I get lastest code form github, and just run "bin/spark-shell --master yarn
--conf spark.dynamicAllocation.enabled=true --conf
spark.dynamicAllocation.initialExecutors=1 --conf
spark.shuffle.service.enabled=true". There is error infor:
15/10/25 12:11:02 ERROR TransportChannelHandler: Connection to
/9.96.1.113:35066 has been quiet for 120000 ms while there are outstanding
requests. Assuming connection is dead; please adjust spark.network.timeout if
this is wrong.
15/10/25 12:11:02 ERROR TransportResponseHandler: Still have 1 requests
outstanding when connection from vm113/9.96.1.113:35066 is closed
15/10/25 12:11:02 WARN NettyRpcEndpointRef: Ignore message
Failure(java.io.IOException: Connection from vm113/9.96.1.113:35066 closed)
15/10/25 12:11:02 ERROR YarnScheduler: Lost executor 1 on vm111: Slave lost
>From log, when driver sends message "GetExecutorLossReason" to AM, the error
>appears. From code, i think AM gets this message, should reply.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
