asterisk users - Sep 2005 - Asterisk 1.0.9 long term stability

I've been evaluating asterisk for quite some time now and am attempting to 
create services on it. The system is simple right now. asterisk seems to 
look up atleast every week if not more. I am running asterisk 1.0.9 and 
would like to find similiar experiences of long term stability.

I attempted to debug it, but my asterisk isn't compiled with all the 
possible debugging flags, which flags in the Makefile should I enable to 
help provide more information? 

Here is what I have found so far.

gdb attach backtrace:
(gdb) bt
#0 0x401c4a76 in nanosleep () from /lib/libc.so.6
#1 0x0000000c in ?? ()
#2 0x401ef4ba in usleep () from /lib/libc.so.6
#3 0x00000000 in ?? ()
#4 0x8a9fa304 in ?? ()
#5 0x8a9ffbe0 in ?? ()
#6 0x00000000 in ?? ()
#7 0x000003e8 in ?? ()
#8 0x8a9fa504 in ?? ()
#9 0x40678bbb in zt_handle_event (ast=0x8a9fa504) at chan_zap.c:590
Previous frame inner to this frame (corrupt stack?)
(gdb) info threads 

This is definitely something in zaptel (zt_handle) but the other errors like 
corrupt stack lead me to believe there is also something else wrong.

Any input would be greatly appreciated.
-- 
Sig Lange
http://www.signuts.net/
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.digium.com/pipermail/asterisk-users/attachments/20050914/c138806c/attachment.htm

Well I don't know how you could measure "long term" stability at
the
moment since 1.0.9 has only been out for about 2 months, but I can offer 
some insight on older versions.

We have one 1.0.3 box that has been up and running for 27+ weeks without 
an issue.  It is running SIP for ~ 30 phones and 1 Cisco gateway.

root@Asterisk-A# asterisk -rx "show uptime"
System uptime: 27 weeks, 1 day, 13 hours, 52 minutes, 57 seconds
Last reload: 3 weeks, 5 days, 18 hours, 29 minutes, 4 seconds

root@Asterisk-A # asterisk -rx "show version"
Asterisk 1.0.3 built by root@Asterisk-A on a i686 running Linux


We have another box that is running 1.0.7 with H.323 to an H.323 
gatekeeper and it is just acting as voicemail for a Cisco Call Manager. 
  It crashes at least 1-2 times per week.  Starting asterisk again 
brings it back up.  I don't know why it happens and I have been unable 
to get anything useful from the logs.  It just dies.

That said, from what I've seen in the past, if you are running SIP, it 
is very stable.  I know I've seen people mention that they have to 
restart it every week or so, but I haven't seen that so far.

Peder


Sig Lange wrote:
> I've been evaluating asterisk for quite some time now and am attempting
> to create services on it. The system is simple right now. asterisk seems 
> to look up atleast every week if not more. I am running asterisk 1.0.9 
> and would like to find similiar experiences of long term stability.
> 
> I attempted to debug it, but my asterisk isn't compiled with all the 
> possible debugging flags, which flags in the Makefile should I enable to 
> help provide more information?
> 
> Here is what I have found so far.
> 
> gdb attach backtrace:
> (gdb) bt
> #0  0x401c4a76 in nanosleep () from /lib/libc.so.6
> #1  0x0000000c in ?? ()
> #2  0x401ef4ba in usleep () from /lib/libc.so.6
> #3  0x00000000 in ?? ()
> #4  0x8a9fa304 in ?? ()
> #5  0x8a9ffbe0 in ?? ()
> #6  0x00000000 in ?? ()
> #7  0x000003e8 in ?? ()
> #8  0x8a9fa504 in ?? ()
> #9  0x40678bbb in zt_handle_event (ast=0x8a9fa504) at chan_zap.c:590
> Previous frame inner to this frame (corrupt stack?)
> (gdb) info threads
> 
> This is definitely something in zaptel (zt_handle) but the other errors 
> like corrupt stack lead me to believe there is also something else wrong.
> 
> Any input would be greatly appreciated.
> -- 
> Sig Lange
> http://www.signuts.net/
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> --Bandwidth and Colocation sponsored by Easynews.com --
> 
> Asterisk-Users mailing list
> Asterisk-Users@lists.digium.com
> http://lists.digium.com/mailman/listinfo/asterisk-users
> To UNSUBSCRIBE or update options visit:
>    http://lists.digium.com/mailman/listinfo/asterisk-users

> We have another box that is running 1.0.7 with H.323 to an H.323
> gatekeeper and it is just acting as voicemail for a Cisco Call  
> Manager.
>   It crashes at least 1-2 times per week.  Starting asterisk again
> brings it back up.  I don't know why it happens and I have been unable
> to get anything useful from the logs.  It just dies.
>
> That said, from what I've seen in the past, if you are running SIP, it
> is very stable.  I know I've seen people mention that they have to
> restart it every week or so, but I haven't seen that so far.
It's not necessarily any specific component of asterisk or version of  
asterisk that can cause instability.  It might have nothing to do  
with asterisk.  One of our boxes is running 1.0.7 (with SIP phones)  
and it has run like a champ for months with no incidents.  I had  
another box that crashed frequently until I was able to come up with  
a good combination of hardware and software versions.

There are so many factors involved.  I read a post on this list not  
too long ago that said the poster has about 30 different asterisk  
boxes running, and they are all identical - asterisk version,  
asterisk patches, OS version, OS patches, telephony hardware, general  
computer hardware, right on down to the bios revision on the  
motherboard.  That was the only way he was able to maintain  
consistent stability across all 30 boxes.  I wish I had time to be  
that organized!

- Noah

lol that was me ironic that I just hijacked this thread and said that
reboots are not a bad thing! It's true I do have 30 IAX/SIP boxen that I
don't reboot, they are all "slave" servers to the IAX/SIP/PRI
"master"
server, which I *do* reboot every night. The 30 boxen I did by cloning a
single hdd and making specific changes to each box. 

this underscores the importance of documenting everything, which I did for
the 30, after I had what I felt was a good config, I documented each single
step to clone the config, getting right down to things like reflashing the
bios, and making sure that a specific brand and rev of nic was in a specific
slot, and that nic used a specific IRQ, and that certain hdw like the
onboard sound card in the chassis was disabled. Armed with such a document,
a third party can come in and easily re-create your work, or troubleshoot
it. The trick is getting the config just right in the first place. As I have
said before, a lot of these issues can be resolved before they happen simply
by not using random hardware and random distro, and to that end, it might
serve Digium well to have a posted HCL and / or sell a certified barebones
box and certify Asterisk on a specific RedHat and / or Debian, and if you
use anything else, well, tough. 

I, myself, would prefer to buy this cerified solution instead of sweating
through a from-scratch config and crossing your fingers and hoping it is
stable. I'm sure a lot of other guys on this list would jump on it as well. 

-----Original Message-----
From: Noah Miller [mailto:noah@rosecompanies.com]
Sent: Wednesday, September 14, 2005 11:00 AM
To: asterisk-users@lists.digium.com
Subject: [Asterisk-Users] Re: Asterisk 1.0.9 long term stability

> We have another box that is running 1.0.7 with H.323 to an H.323
> gatekeeper and it is just acting as voicemail for a Cisco Call  
> Manager.
>   It crashes at least 1-2 times per week.  Starting asterisk again
> brings it back up.  I don't know why it happens and I have been unable
> to get anything useful from the logs.  It just dies.
>
> That said, from what I've seen in the past, if you are running SIP, it
> is very stable.  I know I've seen people mention that they have to
> restart it every week or so, but I haven't seen that so far.
It's not necessarily any specific component of asterisk or version of  
asterisk that can cause instability.  It might have nothing to do  
with asterisk.  One of our boxes is running 1.0.7 (with SIP phones)  
and it has run like a champ for months with no incidents.  I had  
another box that crashed frequently until I was able to come up with  
a good combination of hardware and software versions.

There are so many factors involved.  I read a post on this list not  
too long ago that said the poster has about 30 different asterisk  
boxes running, and they are all identical - asterisk version,  
asterisk patches, OS version, OS patches, telephony hardware, general  
computer hardware, right on down to the bios revision on the  
motherboard.  That was the only way he was able to maintain  
consistent stability across all 30 boxes.  I wish I had time to be  
that organized!

- Noah



_______________________________________________
--Bandwidth and Colocation sponsored by Easynews.com --

Asterisk-Users mailing list
Asterisk-Users@lists.digium.com
http://lists.digium.com/mailman/listinfo/asterisk-users
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users