Not sure. There is a timed restart mechanism, but I'm not
sure if
this is it. I've noticed Caudium do this periodically, but
I've never
dug into why, as it seems to be a clean restart (no core
dumps, etc).
It should be possible to dump all threads when restarting
just to see
what the problem is.
Bill
On Sep 25, 2007, at 8:26 AM, Bertrand LUPART wrote:
> Hello,
>
> Here's what just happened to me.
>
> Top of default.2 file contain this:
>
> ----8<----8<----8<----8<----
> Watchdog not listening on
/tmp/caudium_watchdog_socket_0_26594. Unable
> to turn watchdog on.
> ---->8---->8---->8---->8----
>
> Thus, i expected Caudium not to restart automatically
in case of a
> problem.
>
> Caudium restarted. I don't know why. Nothing in the
file. default.2
> ends
> with:
>
> ----8<----8<----8<----8<----
> ESC[1mCaudium is alive!
> Time: Tue Sep 25 13:01:43 2007
> pid: 13904 ppid: 26595 uid: www-dataESC[0m
> ---->8---->8---->8---->8----
>
> The server restarted automatically.
>
> Did the watchdog his job or is there another restart
mecanism?
|
