[HIVE-14433] refactor LLAP plan cache avoidance and fix issue in merge processor

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.0.1, 2.1.1, 2.2.0
• Fix Version/s: 2.1.1, 2.2.0
• Assignee: Sergey Shelukhin
• Reporter: Sergey Shelukhin
• Created: 04/Aug/16 21:02
• Updated: 08/Dec/16 14:16
• Resolved: 12/Aug/16 20:58

Description

Map and reduce processors do this:

    if (LlapProxy.isDaemon()) {
      cache = new org.apache.hadoop.hive.ql.exec.mr.ObjectCache(); // do not cache plan
...

but merge processor just gets the plan. If it runs in LLAP, it can get a cached plan. Need to move this logic into ObjectCache itself, via a isPlan arg or something. That will also fix this issue for merge processor