I don't see why autorun needs ever know the mud stopped.
Copyover exec's the new mud instance, and exec is
transparent to the parent process.
If you're doing a copyover automatically as part of
SIGSEGV handling (in a fork()'d process) and then
returning so you can dump a core, just do the
copyover from the original process instead of the fork
and autorun will never know.
At 08:05 PM 3/4/99 +0000, you wrote:
>I've noticed with the copyover snippet that Autorun will think the MUD
>died and will keep trying to restart it. Even with the lines:
>
> echo "autorun starting game $DATE" >> syslog
> echo "running bin/circle $FLAGS $PORT" >> syslog
>
> bin/circle $FLAGS $PORT >> syslog 2>&1
>
>#This should catch it...
> while ( /usr/sbin/fuser bin/circle ) do
> sleep 60
> done
>
>
>it still doesn't catch and keeps trying to execute the MUD, causing log
>files to get pushed out of date.
>
>Does anyone know of a way to deal with this?
>
>- Chris Jacobson
+------------------------------------------------------------+
| Ensure that you have read the CircleMUD Mailing List FAQ: |
| http://qsilver.queensu.ca/~fletchra/Circle/list-faq.html |
+------------------------------------------------------------+
This archive was generated by hypermail 2b30 : 12/15/00 PST