Since upgrading to HDP 2.5 I'm getting the following error when running a pig job
2016-11-02 13:57:58,759 [ATS Logger 0] INFO org.apache.hadoop.yarn.client.api.impl.TimelineClientImpl - Exception caught by TimelineClientConnectionRetry, will try 26 more time(s).
If I look under YARN it says the App Timeline Server is running so I can only think that the pig job is looking in the wrong place to talk to it. This wasn't an issue before I upgraded to HDP 2.5
Has anyone got any ideas?
I've noticed the following line higher up in the log file
Obviously the service isn't running on 0.0.0.0 however I can't see where in the config I can define either the hostname or IP address it should be trying to connect to.
Under the YARN config in amabari for the section Application Timeline Server yarn.timeline-service.webapp.address is defined as the "timelineserver.hostname:8188"
I recommend you review the upgrade steps from ATS v. 1.0 to 1.5 document as it involves a lot of manual steps. before trying to configure something on the whim, make sure everything is configured properly https://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.4.2/bk_yarn_resource_mgt/content/ch08s02.html
Did you find resolution to this issue. We are also facing similar issue, but with Tez jobs. We have setup timeline server with correct hostname and it is up and running, but still we see the error:
org.apache.hadoop.yarn.client.api.impl.TimelineClientImpl-Timeline service address: http://0.0.0.0:8188/ws/v1/timeline/
Type pig on command line to verify