On Sunday 18 March 2007 12:57:10 Florent Daigni?re wrote:
> * Dave Baker <dbkr at freenetproject.org> [2007-03-18 12:09:39]:
> 
> > Hi all,
> > 
> > Emu's down again; it's the whole web server this time, rather than just 
HTTPS, 
> > so freenetproject.org is down too.
> 
> Well, that's "normal" since we have migrated everything to apache2
> 
> > 
> > Do we know why this is happenning?
> 
> 'cause a script must be sending a sigterm to apache ... and isn't
> restarting it.
> 
> > It's becoming somewhat of a regular occurrence now: I'm starting to think
> > we should be looking into other hosting arrangements, especially since 
this
> > seems to happen quite a lot on Sundays 
> 
> Well, log rotation scripts are triggered on sundays for instance ;)
> I don't see why it would be our hoster's fault.

Well, okay, when I said 'hosting' I meant the whole server / software / 
connection arrangement, not just the co-lo. The point being that it's looking 
increasingly like *something* should be done, even if it's just upgrades on 
Emu.

> 
> > when none of the people with appropriate karma are around to fix it.
> 
> Excuse us : we have got a life.

I didn't mean to offend. It's just a little akward that this seems to happen 
when nobody is around. Is there some kind of simple solution like adding a 
user group to Emu that only has permission to execute one command that 
restarts the web server? That way you could give shell accounts to reasonably 
trusted people to enable them to log in and prod the web server without 
having to give root to all and sundry.

It's just a bit frustrating to have the downtime when people aren't around, 
and kind of lessens the advantage of having a dedicated server as opposed to 
something like sourceforge.


Dave


> 
> NextGen$
> 
> [Sun Mar 18 06:33:35 2007] [warn] child process 13353 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:35 2007] [warn] child process 18229 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:35 2007] [warn] child process 4067 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:37 2007] [warn] child process 13353 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:37 2007] [warn] child process 18229 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:37 2007] [warn] child process 4067 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:39 2007] [warn] child process 13353 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:39 2007] [warn] child process 18229 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:39 2007] [warn] child process 4067 still did not exit, 
sending a SIGTERM
> [Sun Mar 18 06:33:41 2007] [error] child process 13353 still did not exit, 
sending a SIGKILL
> [Sun Mar 18 06:33:41 2007] [error] child process 18229 still did not exit, 
sending a SIGKILL
> [Sun Mar 18 06:33:41 2007] [error] child process 4067 still did not exit, 
sending a SIGKILL
> [Sun Mar 18 06:33:42 2007] [notice] caught SIGTERM, shutting down
> 
> 



Reply via email to