[
https://issues.apache.org/jira/browse/FALCON-670?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15823059#comment-15823059
]
Mohd Izhar Firdaus Bin Ismail commented on FALCON-670:
------------------------------------------------------
Faced this issue while testing out falcon on HWX sandbox.
I noticed that hive-site.xml is only included if input or output feed is of
table type and is not included otherwise.
ProcessExecutionWorkflowBuilder.java (build(), line 95)
https://github.com/apache/falcon/blob/d55cdd0951bc2c5344ce5e9493c290e58d85234a/oozie/src/main/java/org/apache/falcon/oozie/process/ProcessExecutionWorkflowBuilder.java#L95
IMO Falcon should create that file regardless of input/output feed if the
engine for the Process is hive.
> Hive action in oozie workflow fails on "conf/hive-site.xml" not found
> ---------------------------------------------------------------------
>
> Key: FALCON-670
> URL: https://issues.apache.org/jira/browse/FALCON-670
> Project: Falcon
> Issue Type: Bug
> Components: general
> Affects Versions: 0.5
> Environment: QA
> Reporter: Karishma Gulati
>
> In an oozie workflow, with a specified pig and hive action, the hive action
> fails with this error:
> {code}
> JA009: File does not exist:
> /projects/ivory/staging/falcon/workflows/process/agregator-coord16-022d6601/c1bff9c001d27b2168dbd46f73a5fd2b_1409814409055/user/conf/hive-site.xml
> at
> org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:39)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsUpdateTimes(FSNamesystem.java:1341)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsInt(FSNamesystem.java:1293)
> at org.apache.hadoop.hdfs.server.n
> {code}
> Falcon should be able to get the hcatalog endpoint and generate this, instead
> of the user having to upload the hive-site.xml manually.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
