[OOZIE-2467] Oozie can shutdown itself on long GC pause

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 4.3.0
• Assignee: Purshotam Shah
• Reporter: Purshotam Shah
• Created: 24/Feb/16 19:00
• Updated: 02/Dec/16 21:05
• Resolved: 26/May/16 23:02

Description

There are two issues.
1. Oozie doesn't set ZK session timeout. So, curator uses default session timeout which is 1 min.
If there is GC pause for more than 1 min, or frequent GC pause because of which curator can't send heartbeat, the connection will be dropped and Oozie will shutdown itself.

2. Oozie uses ExponentialBackoffRetry with initial wait = 1sec and max retries = 3.
so Oozie will only retry 3 times, with duration 1sec, 2 sec and 4 sec. We need increase max retries count to at least 10.

Issue Links

• relates to: OOZIE-1847 HA - Oozie servers should shutdown (or go in safe mode) in case of ZK failure