Eugene Koifman created HIVE-18808:
-------------------------------------
Summary: Make compaction more robust when stats update fails
Key: HIVE-18808
URL: https://issues.apache.org/jira/browse/HIVE-18808
Project: Hive
Issue Type: Improvement
Components: Transactions
Affects Versions: 1.0.0
Reporter: Eugene Koifman
Assignee: Eugene Koifman
Worker.gatherStats() runs a "analyze table..." command to update stats which
requires SessionState. SessionState objects are cached in ThreadLocal. If for
some reason Session init fails, it may still get attached to the thread which
then causes a subsequent request that uses the same tread to gather stats fail
because it has a bad session object. HIVE-15658 describes the same issue in a
different context.
There is currently no way to recycle a session from outside HMS.
Failure to gather stats should not kill a compaction job which then prevents
Cleaner from running.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
