I’m running 1.2.30.1 and my server cpu keeps spiking up to 100% and if not restarted goes to 200, then 300, then 400%.
There are only around 40 extensions, asterisk manager is disabled, there are no mp3 files for moh and there are no high throughput queues.
The problem clears after an asterisk restart but then when triggered by something goes immediately back to 100% (not gradually). I’ve installed a cron job which automatically restarts asterisk when the cpu gets > 100% but it is only a sticking plaster, any ideas anyone please?
Hi
What do the logs say was happening when this happened ?
It could be anything realy, looped dialplan, agi who knows what else…
Ian
Hi Thanks for the quick response, I’ve been watching the cpu using ‘top’ and trying to match events to the spike using
tail -f /var/log/asterisk/messages
It isn’t not very effective though, do you know of a better way of logging the possible trigger for the cpu spike?
Looped dial plan is a good one, there are a couple of extensions I’m going to check.
thanks again.
Might be good for you to post the dial plan (Xing out the sensitave materials of course!). Many eyes make light work.
Are you using the flash based operator panel? That thing eats CPU. Also, I remember I had an issue with the voicemail spiking the CPU loads to 100%+ this happened when I reprogrammed the C code to take out some voicemail options…