Bug 182025
| Summary: | mysql logrotate script corner case | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Dave Jones <davej> |
| Component: | mysql | Assignee: | Tom Lane <tgl> |
| Status: | CLOSED RAWHIDE | QA Contact: | David Lawrence <dkl> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 4 | CC: | byte, hhorak, pfrields |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2006-05-18 01:35:26 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
This might be a problem with one of the entries in /etc/logrotate.d/conf or one of the called scripts. indeed. it's mysql.
(05:10:16:davej@firewall:logrotate.d)$ grep kill *
mysqld: [ -e /var/lock/subsys/mysqld ] && /bin/kill -HUP `cat /var/run/mysqld/my
sqld.pid 2> /dev/null ` || /bin/true
mysqld: [ -e /var/lock/subsys/mysqld ] && /bin/kill -HUP `cat /var/run/mysqld/my
sqld.pid 2> /dev/null ` || /bin/true
(05:10:31:davej@firewall:logrotate.d)$ [ -e /var/lock/subsys/mysqld ] &&
/bin/kill -HUP `cat /var/run/mysqld/mysqld.pid 2> /dev/null ` || /bin/true
usage: kill [ -s signal | -p ] [ -a ] pid ...
kill -l [ signal ]
Somehow I ended up with a /var/lock/subsys/mysqld, but no /var/run/mysqld/mysqld.pid
ahh..
Feb 14 14:34:17 firewall kernel: [17700018.240000] Out of Memory: Killed process
2056 (mysqld).
not good. (That was actually a side-effect of tethereal going nuts and leaking
memory -- another bug already filed on that one)
This is an abnormal situation, but the mysql logrotate script should check for
the existence of the pid file before feeding it to kill.
On further review, that logrotate script is unfixably broken anyway; mysqld is simply not designed in a way that allows its stdout/stderr output to be rotated (see also bug #786300). Conclusion is to just remove it, which is now done. I'm sorry, the above reference should be to bug #180639 ... got my bz and it numbers confused ... |
Subject: Cron <root@firewall> run-parts /etc/cron.daily /etc/cron.daily/logrotate: usage: kill [ -s signal | -p ] [ -a ] pid ... kill -l [ signal ] usage: kill [ -s signal | -p ] [ -a ] pid ... kill -l [ signal ] /etc/cron.daily/logrotate looks sane.. #!/bin/sh /usr/sbin/logrotate /etc/logrotate.conf EXITVALUE=$? if [ $EXITVALUE != 0 ]; then /usr/bin/logger -t logrotate "ALERT exited abnormally with [$EXITVALUE]" fi exit 0 logrotate-3.7.1-10