[Bug 326768] Re: mysqld_safe thinks mysqld has crashed when it hasn't

[Mario Limonciello]

> Agreed. It seems that the latter has been identified and a workaround is
> described in this bug. I think the former should also be well understood
> before pushing anything to a *stable* release. A known broken behaviour
> with a workaround is better then pushing an update that breaks existing
> production systems.
I'm sorry, but I entirely disagree here.

Regardless of what is sending a SIGHUP to mysqld_safe, it should be a
supported scenario to allow such signals to be sent to system daemons.
It's common for SIGHUP to be used to ask to reload configuration files
when the daemon supports it.

The "broken" patch from debian's sole purpose is adding support for
catching SIGHUP and a few other signals.  It doesn't work properly.

Ignoring the fact that mysqld is getting restarted rather than reloaded,
the SIGHUP trap support to issue a refresh would *only* work if you
configured /root/my.cnf or had no root mysql password defined in the
first place.

How can reverting a portion of it break an existing production system in
any way?

-- 
mysqld_safe thinks mysqld has crashed when it hasn't
https://bugs.launchpad.net/bugs/326768
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to mysql-dfsg-5.0 in ubuntu.