[MCOL-2132] primproc crash - thread pool issue Created: 2019-02-01  Updated: 2019-02-04  Resolved: 2019-02-04

Status: Closed
Project: MariaDB ColumnStore
Component/s: PrimProc
Affects Version/s: 1.1.5
Fix Version/s: 1.2.2

Type: Bug Priority: Major
Reporter: David Hill (Inactive) Assignee: Unassigned
Resolution: Not a Bug Votes: 0
Labels: None
Environment:

1um 3pm


 Description   

Customer reporting Database locks up on CREATE TABLE IF NOT EXISTS and/or ALTER TABLE x RENAME TO y.

Investigating of issue showed that the PrimProc on all 3 PMs are crashing.

trace captured

[0x561c5a5eb6d0]
/lib64/libpthread.so.0(+0xf6d0)[0x7f04877296d0]
/lib64/libc.so.6(+0x155e68)[0x7f0486875e68]
/usr/local/mariadb/columnstore/lib/libmessageqcpp.so.1(_ZN11messageqcpp10ByteStream6appendEPKhj+0x56)[0x7f0488c52016]
[0x561c5a5a9911]
[0x561c5a5a9bf3]
[0x561c5a5ba259]
/usr/local/mariadb/columnstore/lib/libthreadpool.so.1(_ZN10threadpool18PriorityThreadPool9threadFcnENS0_8PriorityE+0x37b)[0x7f048980a6cb]
/lib64/libboost_thread-mt.so.1.53.0(+0xd27a)[0x7f04881c127a]
/lib64/libpthread.so.0(+0x7e25)[0x7f0487721e25]
/lib64/libc.so.6(clone+0x6d)[0x7f048681ebad]

 Comments   
Comment by David Hill (Inactive) [ 2019-02-04 ]

reported on by development as not a true crash..

i think the crash trace is a red herring, usually is when you see it in messageqcpp. Looking in the logs from pm1 PrimProc memory went above limits and so was killed by ProcMon so that was the cause of the trace.
Jan 31 02:28:55 ip-10-59-10-26 ServerMonitor[9323]: 55.542320 |0|0|0| I 09 CAL0000: Memory Usage for Process: PrimProc : Memory Used 11647115 : % Used 90

Generated at Thu Feb 08 02:33:59 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.