On Fri, Jul 15, 2011 at 4:54 PM, ChristoDeluxe <nab...@cfs.parliant.com> wrote: > Looking in the archives, I see people (mostly back a few years) saying that > people should not use "apachectl restart" or "apachectl graceful" (or > sending the HUP signal) to restart Apache since it leaks memory.
That was the case in the past. Maybe someone else has found a way around it. I haven't. One thing you could try is using graceful shutdown, followed by start. This actually shuts down, but doesn't cut off any existing requests. Another common approach is to have a load balancer that allows you to take half of your cluster offline and restart, and then do the other half. - Perrin
