[Asterisk-Users] Static/Breaking up after IupgradedAsteriskaswell as a crash - Can't trace bug

Mon Jan 10 12:41:53 MST 2005

About 15 of the 20 phones are plugged into a Cisco Powered Switch. About 5
of the phones are using Power Cubes. The specific users having problems,
their phones are powered off the powered switch. The powered switch is
plugged into a UPS. The 24 port powered switch has about 75% phones and 25%
regular computers connected to it, and it is using VLAN's to separate the 4
LAN segments I mentioned earlier. 

Thanks for the suggestion though, I'll try and keep an eye out for customers
reporting quality issues that do use Power Cubes. I noticed there are 2
different types of Power Cubes, official Cisco ones, and aftermarket ones,
which were you using?

  _____  

From: asterisk-users-bounces at lists.digium.com
[mailto:asterisk-users-bounces at lists.digium.com] On Behalf Of Alexander
Lopez
Sent: Monday, January 10, 2005 1:18 PM
To: asterisk-users at lists.digium.com
Subject: Re: [Asterisk-Users] Static/Breaking up after
IupgradedAsteriskaswell as a crash - Can't trace bug

If your remotes are not reporting any trouble. This may be farfetched but
power may be to blame. I have had the ciscos 'freak out' with unstable
power. It looks like the load on the power cubes cannot keep the caps loaded
to deal with fluctuations. Or you may have a ground loop somewhere.

Are the phones plugged into UPSs?

I had flaky 7960 work fine after pluged into a Cisco POE switch.

Keep me in the loop I would like to are how this turns out.

-----Original Message-----
From: asterisk-users-bounces at lists.digium.com
<asterisk-users-bounces at lists.digium.com>
To: 'Asterisk Users Mailing List - Non-Commercial Discussion'
<asterisk-users at lists.digium.com>
Sent: Mon Jan 10 12:49:40 2005
Subject: RE: [Asterisk-Users] Static/Breaking up after I
upgradedAsteriskaswell as a crash - Can't trace bug

Sure. I have yet to test inter-office calls, I'm not sure if it's out to the
PSTN or not. But I would assume if it's Asterisk, then inter-office calls
should suffer as well?  Will have to test that further, as well as phones
connected to Sipura converters, see if they have issues. But every minute
I'm running this version I'm afraid it'll up and crash on me again.

We have 4 PRI's plugged into a Cisco 3640 Router, which then converts an
incoming call to SIP and sends it to the Main Asterisk server. They're
connected via a Cisco Catalyst SmartSwitch (not PoE), but with no VLAN's or
QOS or anything special configured on it. It's really only got the Cisco
Router, the Main Asterisk Server, the Backup Asterisk Server, and our
Companies Asterisk server off of it. An incoming call then goes from the
Main Asterisk Server to our Companies Asterisk server, via IAX and then from
there our Asterisk server sends it to our phones via SIP. It's all local.
There are several of our partner companies registering with the Main Server,
for now I've left that on the 10/26/04 version. No severe quality issues
reported on their end.

Incoming:

PRI -> Cisco 3640 -SIP--> Main_Asterisk (10/26/04) -IAX--> Our_Asterisk
(1/06/05) -SIP--> Cisco 7960 IP Phone

Local Outgoing:

Cisco 7960 -SIP--> Our_Asterisk (1/06/05) -IAX--> Main_Asterisk -SIP-->
Cisco 3640 --> PRI

Long Distance Outgoing:

Cisco 7960 -SIP--> Our_Asterisk (1/06/05) -IAX--> Main_Asterisk -IAX-->
NuFone

At first I thought it was NuFone, but I've heard this reported on local
phone calls as well, and I think (but am not positive) I've heard of this
happening on incoming calls as well.

I think the first thing I need to do is isolate whether inter-office calls
are garbled, this will isolate it to CVS 1/06/05; I think my next step is to
try a Sipura SPA-2000 converter, see if it experiences quality issues
inter-office or out to the PSTN; Are there known issues with IAX linking
between older and newer versions?

I have IAX trunking turned off, since it's all local anyway, but I am using
the ZapRTC hack on our Asterisk server, the Real-Time-Clock replacement to
give me Zaptel timing, so I do have zaptel timing. I have the modules zaprtc
and zaptel loaded, and I have rtcsetup program running in the background on
our Asterisk server. I have used meetme with Our_Asterisk and it works
great.

I had issues with quality when I tried to change the kernel on the
Main_Asterisk server, I tried to upgrade to the latest stock kernel, but
with RTC not compiled in, this way I could also use zaprtc. I'm not why
Asterisk liked the standard Redhat kernel much more than the stock kernel.
Anyways, I don't have zaptel timing working on the Main_Asterisk server, but
since trunking is off and I don't use meetme or anything, it's never been an
issue.

I don't/have not used any extra-curricular Asterisk patches. Most of them
confuse or concern me anyway.

  _____ 

From: asterisk-users-bounces at lists.digium.com
[mailto:asterisk-users-bounces at lists.digium.com] On Behalf Of Alexander
Lopez
Sent: Monday, January 10, 2005 12:13 PM
To: Asterisk Users Mailing List - Non-Commercial Discussion
Subject: RE: [Asterisk-Users] Static/Breaking up after I upgraded
Asteriskaswell as a crash - Can't trace bug

I am also running a pretty recent version albeit not today's CVS, but
CVS-HEAD-11/20/04-11:29:52.

D you have problems b/w the Ciscos or only when going out to the PSTN??

I have 35 7960s with a PRI and no problems that you speak of.  I do get an
occational dropped call but that may be the DHC server lease running out on
the phones.

Can you tell us a little more about your setup???

-----Original Message-----
From: asterisk-users-bounces at lists.digium.com
[mailto:asterisk-users-bounces at lists.digium.com] On Behalf Of Paul Rodan
Sent: Monday, January 10, 2005 11:58 AM
To: 'Asterisk Users Mailing List - Non-Commercial Discussion'
Subject: [Asterisk-Users] Static/Breaking up after I upgraded Asterisk
aswell as a crash - Can't trace bug

We use Cisco 7960's with the P0S3-7-3-00 firmware, which was the latest as
of a few months ago.

I've so far found CVS-v1-0-10/26/04-07:28:01 to be the best version of
Asterisk I've found. I've upgraded regularly in the past, like every other
week. I upgraded to this version and also encountered no issues. About a
month later I tried upgrading, to some version in November, and that's when
all the phones in my office started experiencing quality issues, breaking
up, garbled voice, maybe static. One person reported they had issues
transferring calls, but I could not verify. So I immediately downgraded back
to 10-26-04 and stayed on that version up until several days ago, when I
hoped whatever bug was introduced had been repaired. So I upgraded to
CVS-v1-0-01/06/05-01:53:07 and for a time I thought everything was fine, but
now I get the occasional report of quality issues again, phones breaking
up/garble. It's not as bad as it was before, but I myself have started
experiencing quality issues on my phone and I've never experienced these
issues before. So whatever bug existed, still exists. It might not be a bug,
but maybe some modification to chan_sip has broken compatibility with the
Cisco 79xx phones. Unfortunately I am not a developer and am not able to
take apart and put back together again the source, to adapt it to my own
needs, so I'm at the mercy of the Asterisk developers.

Nothing has changed in our network topology, no new phones added, no
computers share the subnets with the phones. They're part of the same
physical network as our computers, but do have their own separate subnet. I
haven't tried other phones, or converters, but I can if anybody wants me to
do further analysis. What I have here is a unique situation to single out a
quality issue, a bug, and I'd like to help by testing different versions of
chan_sip.c to see which option/modification in fact created the quality
issue. I would stay on this version of Asterisk longer (despite the
occasional quality issues) if it weren't for the fact that yesterday evening
the Asterisk daemon crashed for no apparent reason. Until this time, the 6
months I've been running Asterisk, Asterisk has never crashed on me. All the
phones in one of my Call Groups started ringing for no reason, when I
answered, nobody was there. I had to go and answer each phone individually,
there wasn't anybody on any phone. After the last was answered and hung up,
the phones were quiet. But when I tried to access voicemail or dial out,
nothing, that's how I'd know Asterisk had crashed, the
/var/log/asterisk/messages file revealed nothing, absolutely nothing,
neither did /var/log/messages

So tonight I'm going back down to 10-26 and I'd bet money the quality issues
disappear, it's happened before. I hate downgrading, I feel like I'm now
stuck at a certain version and am unable to proceed safely. The security and
bug fixes I keep seeing hit the stable version are all now no longer
available to me, which sucks. Can anybody suggest how I can trace this
issue? During one of the phones conference calls the quality was really
horrible, so I started a constant ping on the phone to see if there were
jumps in latency or packet loss, as their very sensitive to this, and I
didn't see any.

The only other thing I can think of is bad phones? One sales guy had
continuous phone quality issues since the upgrade, so I traded my phone for
his, switched the config files. And I made his mine. And when I started
placing calls, I started having quality issues in which I didn't have any
before. So I thought the phone must be defective. So then I grabbed another
phone, a Cisco 7940 and made it mine, and the quality is better, but I still
here the occasional robotic sound when I place calls, and this is a
completely different phone. I'm thinking I just didn't notice the problems
as severely as others have. There's also 4 or 5 other employees (about 20 of
us in total) reporting quality issues, I don't think it's possible for that
many phones to fail when they've been doing good for so long.

Any help or advice would be helpful. However, a couple of my friends
companies run asterisk and I've already seen the "I use CVS Version so and
so (newer than mine) and I don't have any problems, etc. etc." lines, so I
know that for most it must work, but I'm one of the ones that it does not
work for, and I'd really be interested in finding out why. I've pretty much
eliminated the network possibility. They're all local, we have 4 physical
segments each with 2 subnets, 1 for the computer and 1 for the phones, the
subnets are all linked by a Linux firewall with multiple interfaces. This
firewall is set to give the VOIP Subnets full access to one another, and has
never interfered with the phones ability to communicate with the Asterisk
server before.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20050110/de02ec32/attachment.htm