Pig jobs from pig view fail with unexcepted end of file

[#1]

Pig jobs work from gateway node but fail from ambari-view. I crosschecked configs with proxyusers and I have ambariid which is running ambari-server configured there.

Error from RM shows this.

AM Container for appattempt_1463770749228_0048_000001 exited with exitCode: -1000 For more detailed output, check application tracking page:http://<hostname>:8088/cluster/app/application_1463770749228_0048Then, click on links to logs of each attempt. Diagnostics: ExitCodeException exitCode=2: gzip: /grid/6/hadoop/yarn/local/filecache/33_tmp/tmp_pig.tar.gz: unexpected end of file tar: Unexpected EOF in archive tar: Unexpected EOF in archive tar: Error is not recoverable: exiting now Failing this attempt

Error from Pig view is File does not exist: /user/rmutyal/pig/jobs/test_23-05-2016-14-46-54/stdout but webhcat user configuration looks alright. What am I missing?

[#2]

@Ravi Mutyala

Please check HTTP Proxy. Change them as below in core-site.xml and restart affected services:

hadoop.proxyuser.HTTP.groups = *

hadoop.proxyuser.HTTP.hosts = *

Let me know if you are able to execute. It worked for me.

[#3]

No luck with that. This is a cluster with https configured for ambari

[#4 · Accepted solution]

This is a case of corrupt pig.tar.gz in hdfs (/hdp/apps/<version>/pig) folder. I am not sure how it ended up with a corrupt version there on a fresh install based on ambari. But once I manually updated with pig.tar.gz from /usr/hdp/<version>/pig/, the error got resolved.

However, confusing part is pig view throwing a completed unrelated error (File does not exist at /user/rmutyal/pig/jobs/test_23-05-2016-14-46-54/stdout)