[IMPALA-1043] Instability running queries on a cluster with low mem limit

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: Impala 1.4
• Fix Version/s: Impala 1.4
• Assignee: Nong Li
• Reporter: Nong Li
• Created: 12/Jun/14 01:45
• Updated: 14/Jun/14 03:15
• Resolved: 14/Jun/14 03:15

Description

I ran one of the tpcds queries with a very low me limit which makes it cancel every time. After about 10K runs, it caused some of the nodes to fail:

The failed nodes were
c2308.halxg.cloudera.com: Core dumps found!
c2334.halxg.cloudera.com: Core dumps found!
c2336.halxg.cloudera.com: Core dumps found!

and the core dumps are at /data/5/coredumps/2507/.

Each node has a different stack and it looks like some objects were deleted while being used.