Aastra phones dropping to No Service
Submitted by alucky8ball on Wed, 04/29/2009 - 7:08am.
Hi all,
I am having a problem on a trixbox setup with approximately 40 Aastra phones where the phones will drop to no service, then a few moments later go back to normal. When I login to the web interface of the phone, it shows a status of "408". Any ideas what could be causing this or any logs that I could post to help?
Thanks,
Jeremy
 
Aastra Phones
Wed, 04/29/2009 - 10:52am
Software load? We have had problems with newer software loads. Most reliable has been 2.2.1.25
 
We've got the same issue -
Wed, 04/29/2009 - 1:52pm
We've got the same issue - hopefully we can get the error code from the Web interface tomorrow.
It's a newer software build too... can Aastra's be downgraded?
 
Network issue maybe ?
Thu, 04/30/2009 - 6:52am
I have seen the same issue on my aastra phone. Tried a wireshark trace and noticed that phone was sending re-REGISTERs but there was no response from the server. Eventually phone gets 408 request timeout message. Then phone kept sending re-REGISTERs every 120 seconds and server responded with 200ok. Phones came back into service. Yet to know if this is a network issue or trixbox but phone seemed to do the correct call flow. Have you guys tried running wireshark trace yet? might help
 
I originally thought network
Thu, 04/30/2009 - 7:08am
I originally thought network issue. We then spent several days trying to find a problem in the network, upgraded the IOS on all of our switches (Cisco Catalyst 3750-48PS). However, all tests on the network came back perfectly fine. I have noticed that if I reboot the trixbox system, then the no service messages aren't as frequent. I'm leaning towards something in TB, but not sure exactly what it is. I did notice that directly after a reboot the memory usage is only about 80% on my box, then after about a day or so it spikes up to high 90's.
Jeremy Rich - FTOCC Boston
Eastern Carolina Vocational Center
 
Are you doing a lot of
Thu, 04/30/2009 - 1:17pm
Are you doing a lot of paging to numerous extensions?
 
We are at this site; there
Sat, 05/09/2009 - 2:16am
We are at this site; there are quite a few pages every day. Paging 50+ extensions
How would this affect it?
 
I finally gave in and called
Mon, 05/11/2009 - 1:22pm
I finally gave in and called Trixbox Support. The issue for me ended up being that the DNS server was sometimes unreachable (Due to our ISP). The best I could tell, there is a bug in Asterisk that will drop all SIP registrations if it cannot communicate with DNS. I ended up changing the DNS to an inside host and things seem to be much happier.
Jeremy Rich - FTOCC Boston
Eastern Carolina Vocational Center
 
I'm having a similar problem
Thu, 07/02/2009 - 5:24am
I'm having a similar problem with 2.6.2.2
I can knock the phones into "No service" by copying a directory from one place to another on the same hard drive. The directory I'm copying was bout 650MB. About halfway through the copy, all of my phones drop service. My last install was 2.2.3 and I never had this particular issue.
I guess I'm going to need to sniff the network to see if I can see anything happening.
I've got my DNS pointing to an internal server that's available so I'm not sure that's the cause, but I've been doing phone to server resolution by IP anyway so I fail to see where DNS would become an issue.
One thing I noticed today is that if I tried to call an extension, say 140, while the service was down, the phone would try to "call" the phone with IP address 10.1.2.140. The phones and the trixbox are on the 10.1.2.x subnet.
If I dialed a phone with a known IP address, like 10.1.2.200, I could successfully make an interoffice call by dialing 200 even though it's actual extension was 125 for example.
 
Having same issue with
Wed, 07/08/2009 - 8:35pm
Having same issue with 2.6.2.2 with 5 Aastra phones and also with 2.6.1.11 with 3 Aastra phones. However a client with 2.6.1.10 has 50 Aastra phones and no issue. The latter is using OPEN DNS and the phones use the DHCP server in trixbox. The 2.6.2.2 uses WIN 2003 server DNS and the phones use WIN 2003 DHCP. The 2.6.1.11 uses Untangle DNS and Cisco DHCP.
I've noticed that on the 2.6.2.2 if I try to open a large text file (3G) from /var/log/asterisk that 3 of the 5 phones will drop registration and then half a minute later come back up. The 2.6.1.11 drops randomly for no apparent reason.
Also the cpu usage rises from 0 to about 30% to over 60% for about a 1.75 minutes and then to 100% over about a 2 minute period that it takes to have "less /var/log/asterisk/full.2" open the 3GB asterisk log file. (This file was made by a beta BRI driver for the Sangoma card.) So in this case anyway it seems to be that when the processor has a heavy load that the phones are dropped.
 
same problem with PIAF, so Idon't think it's a trixbox issue.
Sun, 07/26/2009 - 8:22pm
Testing with 2 Aastra phones and emitting the same behavior. Mine doesn't come back sometimes, and rebooting the phones and server sometimes brings them back.
 
Member Since:
2007-06-06