[Spce-user] production system - not accepting registration
Matthew Ogden
matthew at tenacit.net
Tue Apr 23 05:40:07 EDT 2013
Hi Andrew,
No, sorry, it occurred this morning (the first log I sent I had used the
wrong day in grep). (I rebooted it at 10:40 today to make it work again).
Memory usage seems a lot higher in 2.8 than 2.5, is that expected? Will
upgrade it today, anything I can do to keep memory usage down till then?
I think it might be related to when doing queries on CDRs, its seems much
more intensive with the "details" section etc?
Kind Regards
> -----Original Message-----
> From: Andrew Pogrebennyk [mailto:apogrebennyk at sipwise.com]
> Sent: 23 April 2013 11:22 AM
> To: Matthew Ogden
> Cc: spce-user at lists.sipwise.com
> Subject: Re: [Spce-user] production system - not accepting registration
>
> Hi,
> on the other hand I see the following lines:
>
> Apr 23 10:12:13 spce kernel: [46951.246967] Out of memory: kill process
> 2128 (kamailio) score 375329 or a child
> Apr 23 10:12:13 spce kernel: [46951.247186] Killed process 2134
> (kamailio) Apr 23 10:12:13 spce kernel: [46951.423903] kamailio invoked
> oom-killer:
> gfp_mask=0x200da, order=0, oom_adj=0
>
> This means the server ran out of memory and OS killed kamailio.
> But this is one hour ago or so, AFTER the server has been rebooted? And
> the issue occurred yesterday morning? In that case you need to check
e.g.
> /var/log/syslog.1.
> Also grep kamailio-lb.log of that day for ERROR string.
More information about the Spce-user
mailing list