This is a place where I can describe my use of Asterisk, and note any incidents along the way.
3-December-2005: I messed with the QoS settings on the router (changed from TCP and UDP for IAX QoS to just UDP) then tried to place a softphone and a hardphone call. Both were not registering. I set SIP debug to on, and reset the Sipura box and restarted the softphone. Nothing was registering with Asterisk. The Sipura was showing that it was unable to register, and the softphone was stuck at the “logging in…” step.
After a few reloads and a sanity check of the configuration files I rebooted the BSD box and everything came back up.
I’ll have to check some of the pages on reliability.
8-December-2006: Found Asterisk running at close to 100% CPU:
40 processes: 2 running, 38 sleeping
CPU states: 62.3% user, 0.0% nice, 37.7% system, 0.0% interrupt, 0.0% idle
Mem: 18M Active, 252M Inact, 85M Wired, 160K Cache, 57M Buf, 108M Free
Swap: 935M Total, 935M Free
PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
42942 root 132 0 3536K 2528K RUN 17.4H 94.78% 94.78% asterisk
240 root 96 0 1788K 1336K select 0:40 0.00% 0.00% dhclient
435 root 96 0 3384K 2532K select 0:09 0.00% 0.00% sshd
452 root 8 0 1364K 1052K nanslp 0:03 0.00% 0.00% cron
I can still call and leave messages, but 100% CPU doesn’t look healthy. Connected with
asterisk -vvvvvr and issued a
stop gracefully and now asterisk is at 98% CPU and
asterisk -vvvvvr connections are getting a
Unable to connect to remote asterisk response.
sudo kill 42942 did the trick. Restarted Asterisk and everything is sane again.
I have read some recommendations that Asterisk be restarted occasionally. That wouldn’t help with availability and reliability metrics at all, but might be ok for my purposes. I’ll see if I can dig up the reference.