On Mon, 8 Jul 2002, Nico Erfurth wrote:
> Philip Hazel wrote:
>
> >>And pkill isn't available anywhere, ok so forget about this, it's too
> >>dangerous.
> >
> > Hmm. I could build something that required a configuration setting. Then
> > the OS-specific files could set killall or pkill for those systems where
> > it was known to work. It seems a shame not to use a faster method where
> > it is available.
>
> Good idea, it's really bad too see so many "No such PID"-messages on a
> high loaded system.
> And the ps/egrep/kill solution COULD send signals to the wrong
> processes, in strange situations. It's very unlikely, but not
> impossible, that another process with the same PID spawns, after the
> exim-process exited.
I wonder hoe difficult it would be to have _every_ exim process write
its pid to a unique file when it starts, and remove it when it exits.
eg
/var/run/exim-4098
/var/run/exim-4104
(the content of the files would have the PID as well)
This would make it easy to get all PID's of all running exims...
>
> ciao
>
>
> --
>
> ## List details at http://www.exim.org/mailman/listinfo/exim-users Exim details at http://www.exim.org/ ##
>
>
