[Asterisk-Users] safe_asterisk not working?

[Adrian A]

I've been experiencing some crashes in Asterisk in the past few weeks.  I
haven't been able to find out why as gdb shows it's in a different function
every time.  But, in the meantime, I've been using safe_asterisk hoping that
it would simply restart Asterisk by itself.  It doesn't seem to do that.
Whenever Asterisk crashed, the list of processes doesn't show asterisk or
safe_asterisk running anymore. I do not get an e-mail notification and the
core dumped is in the standard format e.g. core.18875 not
core.`hostname`-`date -Iseconds`.  Does anyone know what I could do to
troubleshoot this since I can't really force a crash to see what the script
is doing?  I have the following variables set in safe_asterisk:

#!/bin/sh
CLIARGS="$*"            # Grab any args passed to safe_asterisk
TTY=9                   # TTY (if you want one) for Asterisk to run on
CONSOLE=no              # Whether or not you want a console
NOTIFY=myemail at mydomain.com   # Who to notify about crashes
DUMPDROP=/tmp

I also noticed the line
"elif [ $EXITSTATUS -gt 128 ];"
inside the script yet my cores show:
"Core was generated by `asterisk -vvvg'.
Program terminated with signal 11, Segmentation fault."
so perhaps that is why the restart is not triggered?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20060112/1506c5b3/attachment.htm