| Bug #40973 | Server stalls in concurrency test (pushbuild2), metadata locks | ||
|---|---|---|---|
| Submitted: | 24 Nov 2008 11:22 | Modified: | 29 Jul 2009 10:09 |
| Reporter: | Guilhem Bichot | Email Updates: | |
| Status: | Duplicate | Impact on me: | |
| Category: | MySQL Server | Severity: | S3 (Non-critical) |
| Version: | 6.0-bzr,5.4 | OS: | Linux |
| Assigned to: | Assigned Account | CPU Architecture: | Any |
| Tags: | mdl | ||
[2 Jul 2009 8:37]
Konstantin Osipov
Alik, once manage to repeat the bug, let's discuss.
[2 Jul 2009 19:54]
Alexander Nozdrin
The bug has been reproduced.
[16 Jul 2009 9:27]
Alexander Nozdrin
It looks like this is a duplicate of Bug#43867.
Description: Initially this was seen in the 6.0-maria tree (see URL for stack traces of 30 threads; in short there are several threads in mdl.cc, and one in release_table_share()). But running the same stress test on MyISAM in 6.0-main gives a stall too (I got this on two Linux machines, unfortunately gdb doesn't give good traces there, so I cannot provide them and cannot be 100% sure that it's the same stall - but anyway it's a stall). My 6.0-main is [INS 12:05 /m/bzrrepos/mysql-6.0 $] bzr log -r-1 --show-ids --short 2919 Alexander Nozdrin 2008-11-21 revision-id:alik@sun.com-20081121110106-cdkd50bjnqoitlou Revert default.conf How to repeat: get the two files attached to BUG#40954, and run the script as explained in that bug report, but against a 6.0-main tree. I get: # 12:09:41 Started periodic reporting process... # 12:14:43 10 stalled queries detected, declaring deadlock at DSN dbi:mysql:host=127.0.0.1:port=19306:user=root:database=test.