Michael,
Thanks for those ideas. I'm going to wait and see what the next crash
brings. Hopefully a useful core file. If not, then I'll try out your
recommendations from there. Using screen is a good idea :)
Tom.
On Wed, 2004-01-07 at 18:49, Michael Haardt wrote:
> > > Run exim inside a wrapper script to capture any output to stderr and stdout
> > > when it does. RestartWrapper from open system consultants is one that comes
> > > to mind
>
> In case that does not help and the machine is not under heavy load, then
> attach strace to the running daemon, piping its output to tail -1000
> and saving that in a file. It causes way more load, but that way you
> know the last 1000 system calls before the crash, no matter how long it
> stays up, and it won't fill your disk.
>
> Another option is to attach gdb with its input from a file. Ignore
> SIGPIPE in gdb, attach to the process and output a backtrace upon
> returning from the attached process. Yet another option would be running
> the gdb session under screen(1).
>
> Michael
>
> --
>
> ## List details at http://www.exim.org/mailman/listinfo/exim-users Exim details at http://www.exim.org/ ##
>
