Bug #27147 Crash on Query
Submitted: 14 Mar 2007 18:31 Modified: 15 Apr 2007 16:24
Reporter: Karsten Eichentopf Email Updates:
Status: No Feedback Impact on me:
None 
Category:MySQL Server: General Severity:S3 (Non-critical)
Version:5.0.32-Dotdeb_1.dotdeb.1-log OS:Linux (Linux - x86_64 - Debian Sarge)
Assigned to: CPU Architecture:Any
Tags: 5.0.32-Dotdeb_1.dotdeb.1-log, jce, joomla, query, sarge

[14 Mar 2007 18:31] Karsten Eichentopf
Description:
I have two debian sarge root servers running this dotdeb package. After the upgrade to 5.0.32 one query that is executet by the CMS Joomla! and the Module JCE will crash the MySQL Server on both servers. Our other two root servers running standard sarge mysql 4 and etch MySQL 5 show no similar problems.

# Server Version: 5.0.32-Dotdeb_1.dotdeb.1-log
# Protokoll-Version: 10
# Server: Localhost via UNIX socket
# MySQL-Client-Version: 4.1.15

#similar error reported. in german
http://www.joomlaportal.de/joomla-mambots/83357-absturz-des-mysql-servers-mit-jce-editor.h...

Syslog:

Mar 14 18:25:03 prov02 mysqld[2559]: mysqld got signal 11;
Mar 14 18:25:03 prov02 mysqld[2559]: This could be because you hit a bug. It is also possible that this binary
Mar 14 18:25:03 prov02 mysqld[2559]: or one of the libraries it was linked against is corrupt, improperly built,
Mar 14 18:25:03 prov02 mysqld[2559]: or misconfigured. This error can also be caused by malfunctioning hardware.
Mar 14 18:25:03 prov02 mysqld[2559]: We will try our best to scrape up some info that will hopefully help diagnose
Mar 14 18:25:03 prov02 mysqld[2559]: the problem, but since we have already crashed, something is definitely wrong
Mar 14 18:25:03 prov02 mysqld[2559]: and this may fail.
Mar 14 18:25:03 prov02 mysqld[2559]:
Mar 14 18:25:03 prov02 mysqld[2559]: key_buffer_size=16777216
Mar 14 18:25:03 prov02 mysqld[2559]: read_buffer_size=131072
Mar 14 18:25:03 prov02 mysqld[2559]: max_used_connections=2
Mar 14 18:25:03 prov02 mysqld[2559]: max_connections=100
Mar 14 18:25:03 prov02 mysqld[2559]: threads_connected=1
Mar 14 18:25:03 prov02 mysqld[2559]: It is possible that mysqld could use up to
Mar 14 18:25:03 prov02 mysqld[2559]: key_buffer_size + (read_buffer_size + sort_buffer_size)*max_connections = 233983 K
Mar 14 18:25:03 prov02 mysqld[2559]: bytes of memory
Mar 14 18:25:03 prov02 mysqld[2559]: Hope that's ok; if not, decrease some variables in the equation.
Mar 14 18:25:03 prov02 mysqld[2559]:
Mar 14 18:25:03 prov02 mysqld[2559]: thd=0x11e6e00
Mar 14 18:25:03 prov02 mysqld[2559]: Attempting backtrace. You can use the following information to find out
Mar 14 18:25:03 prov02 mysqld[2559]: where mysqld died. If you see no messages after this, something went
Mar 14 18:25:03 prov02 mysqld[2559]: terribly wrong...
Mar 14 18:25:03 prov02 mysqld[2559]: Cannot determine thread, fp=0x4403f768, backtrace may not be correct.
Mar 14 18:25:03 prov02 mysqld[2559]: Stack range sanity check OK, backtrace follows:
Mar 14 18:25:03 prov02 mysqld[2559]: (nil)
Mar 14 18:25:03 prov02 mysqld[2559]: New value of fp=0x11e6e00 failed sanity check, terminating stack trace!
Mar 14 18:25:03 prov02 mysqld[2559]: Please read http://dev.mysql.com/doc/mysql/en/us...ack-trace.html and follow instructions on how to resolve the stack trace. Resolved
Mar 14 18:25:03 prov02 mysqld[2559]: stack trace is much more helpful in diagnosing the problem, so please do
Mar 14 18:25:03 prov02 mysqld[2559]: resolve it
Mar 14 18:25:03 prov02 mysqld[2559]: Trying to get some variables.
Mar 14 18:25:03 prov02 mysqld[2559]: Some pointers may be invalid and cause the dump to abort...
Mar 14 18:25:03 prov02 mysqld[2559]: thd->query at 0x1257f90 = SELECT m.*, u.name AS editor
Mar 14 18:25:03 prov02 mysqld[2559]: FROM jos_jce_plugins AS m
Mar 14 18:25:03 prov02 mysqld[2559]: LEFT JOIN jos_users AS u ON u.id = m.checked_out
Mar 14 18:25:03 prov02 mysqld[2559]: WHERE m.client_id = '0'
Mar 14 18:25:03 prov02 mysqld[2559]: GROUP BY m.id
Mar 14 18:25:03 prov02 mysqld[2559]: ORDER BY m.row ASC, m.ordering ASC, m.type ASC, m.name ASC
Mar 14 18:25:03 prov02 mysqld[2559]: LIMIT 0, 30
Mar 14 18:25:03 prov02 mysqld[2559]: thd->thread_id=16
Mar 14 18:25:03 prov02 mysqld[2559]: The manual page at http://www.mysql.com/doc/en/Crashing.html contains
Mar 14 18:25:03 prov02 mysqld[2559]: information that should help you find out what is causing the crash.
Mar 14 18:25:03 prov02 mysqld_safe[2586]: Number of processes running now: 0
Mar 14 18:25:03 prov02 mysqld_safe[2588]: restarted
Mar 14 18:25:03 prov02 mysqld[2591]: 070314 18:25:03 InnoDB: Started; log sequence number 0 43665
Mar 14 18:25:03 prov02 mysqld[2591]: 070314 18:25:03 [Note] Recovering after a crash using /var/log/mysql/mysql-bin
Mar 14 18:25:03 prov02 mysqld[2591]: 070314 18:25:03 [Note] Starting crash recovery...
Mar 14 18:25:03 prov02 mysqld[2591]: 070314 18:25:03 [Note] Crash recovery finished.
Mar 14 18:25:03 prov02 mysqld[2591]: 070314 18:25:03 [Note] /usr/sbin/mysqld: ready for connections.
Mar 14 18:25:03 prov02 mysqld[2591]: Version: '5.0.32-Dotdeb_1.dotdeb.1-log' socket: '/var/run/mysqld/mysqld.sock' port: 3306 Dotdeb Sarge backport

How to repeat:
Execute the query against the tables from the dump with phpMyAdmin or anything and the crash will happen. If i change the query a bit no crash will happen. If i delete the first 30 lines from jos_jce_plugins no crash will happen. But on my two debian sarge with the dotdeb package and the data given by me i can reproduce the crash any time.

------------------------

SELECT m.*, u.name AS editor FROM jos_jce_plugins AS m LEFT JOIN jos_users AS u ON u.id = m.checked_out WHERE m.client_id = '0' GROUP BY m.id ORDER BY m.row ASC, m.ordering ASC, m.type ASC, m.name ASC LIMIT 0, 30
[14 Mar 2007 18:33] Karsten Eichentopf
Dump of table ised in the query

Attachment: table1.sql (text/x-sql), 6.77 KiB.

[14 Mar 2007 18:33] Karsten Eichentopf
Dump of table ised in the query

Attachment: table2.sql (text/x-sql), 1.12 KiB.

[15 Mar 2007 16:24] Valeriy Kravchuk
Thank you for a problem report. Sorry, but I was not able to repeat the behaviour described with latest 5.0.38-BK. So, please, try to repeat with a newer version, 5.0.37 (MySQL's binaries), and inform about the results.
[15 Apr 2007 23:00] Bugs System
No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".