Anyone have any thoughts on why versions of asterisk I try (4 so far) after CVS-07/18/03 always end up locking up on me... which means no sip clients can register/re-register and if I type "reload" or "stop now" at the cli it just returns and does nothing. I have experienced this same issue on three separate boxes. Two running RedHat 9 and one running Redhat 8. I don't have any digium cards installed. I use SIP only. I use cisco IP phones and cisco sip voice gateways. I have experienced the issue while using ztdummy and zaprtc as well as neither. The lockup occurs once every 2 hours or so during heavy use, which is like 50 calls an hour and people checking their voicemail and such. During off hours like the weekend it usually doesn't even crash at all even though we do have some call volume. probably like 20 calls all weekend. CLI commands such as "sip show peers", "show channels", "show modules", still return results. Thanks, David Harris
Are you using cdr_mysql module ? (storing CDRs in mysql ?) regards Martin On Thu, 28 Aug 2003, David Harris wrote:> Anyone have any thoughts on why versions of asterisk I try (4 so far) > after CVS-07/18/03 always end up locking up on me... which means no sip > clients can register/re-register and if I type "reload" or "stop now" at > the cli it just returns and does nothing. > > I have experienced this same issue on three separate boxes. Two running > RedHat 9 and one running Redhat 8. > > I don't have any digium cards installed. I use SIP only. I use cisco > IP phones and cisco sip voice gateways. I have experienced the issue > while using ztdummy and zaprtc as well as neither. > > The lockup occurs once every 2 hours or so during heavy use, which is > like 50 calls an hour and people checking their voicemail and such. > During off hours like the weekend it usually doesn't even crash at all > even though we do have some call volume. probably like 20 calls all > weekend. > > CLI commands such as "sip show peers", "show channels", "show modules", > still return results. > > Thanks, > > David Harris > > _______________________________________________ > Asterisk-Users mailing list > Asterisk-Users@lists.digium.com > http://lists.digium.com/mailman/listinfo/asterisk-users >
Hi David- As mentioned, there was a bug in the CDR SQL module that caused a problem similar to what you describe, however the symptom I experienced in this case was that the "reload" command would start, but then hang up during the reload of one the modules. I was still able to "stop now" correctly however. I'm also Running RH9. A recent 8/26? CVS replaced cdr_sql, and I think it fixed that particular problem. I have noticed sometimes that asterisk -r fails to connect to a running asterisk server after a several hour period of time - I am trying to get more data on this before reporting to the bug list. Maybe this is related to your problem? Cheers Scott Scott M. Stingel Emerging Voice Technology Inc. Palo Alto, California and London, England Email: scott@evtmedia.com URL: www.evtmedia.com -----Original Message----- From: asterisk-users-admin@lists.digium.com [mailto:asterisk-users-admin@lists.digium.com] On Behalf Of David Harris Sent: Thursday, August 28, 2003 9:09 AM To: asterisk-users@lists.digium.com Subject: [Asterisk-Users] Asterisk stops responding Anyone have any thoughts on why versions of asterisk I try (4 so far) after CVS-07/18/03 always end up locking up on me... which means no sip clients can register/re-register and if I type "reload" or "stop now" at the cli it just returns and does nothing. I have experienced this same issue on three separate boxes. Two running RedHat 9 and one running Redhat 8. I don't have any digium cards installed. I use SIP only. I use cisco IP phones and cisco sip voice gateways. I have experienced the issue while using ztdummy and zaprtc as well as neither. The lockup occurs once every 2 hours or so during heavy use, which is like 50 calls an hour and people checking their voicemail and such. During off hours like the weekend it usually doesn't even crash at all even though we do have some call volume. probably like 20 calls all weekend. CLI commands such as "sip show peers", "show channels", "show modules", still return results. Thanks, David Harris _______________________________________________ Asterisk-Users mailing list Asterisk-Users@lists.digium.com http://lists.digium.com/mailman/listinfo/asterisk-users
Hi, We have experienced the exact same problem. Pure SIP environment and yes...we store CDRs in MySQL. Version is: CVS-08/22/03 Gazing at the console I was able to determine the exact time Asterisk froze. Even with DEBGUG on it did not show anything important. The moment it freezes is when a call from Phone1 tries to connect to a SIP Provider like Iconnect: Phone1----Our SIP Server-------Our Asterisk--------SIP Provider It was by no means 100% reproducible. Maybe 1 out of 10 calls caused the trouble. A bad symptom would be that the command "show sip channels" would show several calls, even though they had hungup a long time ago. Troubleshooting revealed that the BYE message was not being sent by our SIP Server to the Asterisk server upon hangup. We rectified this and we no longer see those phantom SIP Channels and Aterisk has not froze for about a week. Regards, Andres http://www.telesip.net On Thursday 28 August 2003 11:08, David Harris wrote:> Anyone have any thoughts on why versions of asterisk I try (4 so far) > after CVS-07/18/03 always end up locking up on me... which means no sip > clients can register/re-register and if I type "reload" or "stop now" at > the cli it just returns and does nothing. > > I have experienced this same issue on three separate boxes. Two running > RedHat 9 and one running Redhat 8. > > I don't have any digium cards installed. I use SIP only. I use cisco > IP phones and cisco sip voice gateways. I have experienced the issue > while using ztdummy and zaprtc as well as neither. > > The lockup occurs once every 2 hours or so during heavy use, which is > like 50 calls an hour and people checking their voicemail and such. > During off hours like the weekend it usually doesn't even crash at all > even though we do have some call volume. probably like 20 calls all > weekend. > > CLI commands such as "sip show peers", "show channels", "show modules", > still return results. > > Thanks, > > David Harris > > _______________________________________________ > Asterisk-Users mailing list > Asterisk-Users@lists.digium.com > http://lists.digium.com/mailman/listinfo/asterisk-users
>Are you using cdr_mysql module ? (storing CDRs in mysql ?)>regards >MartinNo I am using cdr_csv And a "show modules" confirms this. David Harris
>Gazing at the console I was able to determine the exact time Asterisk >froze. >Even with DEBGUG on it did not show anything important. The moment it >freezes is when a call from Phone1 tries to connect to a SIP Providerlike>Iconnect:I have not been able to pin point exactly what event causes the freeze-up but I have been on the console when it has happened. It didn't print out anything interesting. The call I was on cut off.>Phone1----Our SIP Server-------Our Asterisk--------SIP Provider>It was by no means 100% reproducible. Maybe 1 out of 10 calls causedthe>trouble.Same here except I would say more like 1 out of 100 calls.> A bad symptom would be that the command "show sip channels" >would show several calls, even though they had hungup a long time ago.I definitely have this problem.>Troubleshooting revealed that the BYE message was not being sent by ourSIP>Server to the Asterisk server upon hangup. We rectified this and we no >longer see those phantom SIP Channels and Aterisk has not froze forabout a >week. What is your "SIP Server" what does it do? Maybe I have the same issue with my Cisco Voice Gateway not sending the BYE message sometimes. But would this cause asterisk to freeze? Other "symptoms" I have are these errors in the asterisk messages log file Aug 27 09:21:00 NOTICE[1081364]: File sched.c, Line 209 (sched_settime): Request to schedule in the past?!?! Aug 27 09:21:24 NOTICE[1081364]: File sched.c, Line 209 (sched_settime): Request to schedule in the past?!?! Aug 27 09:21:29 NOTICE[1081364]: File sched.c, Line 209 (sched_settime): Request to schedule in the past?!?! Aug 27 09:21:35 NOTICE[1081364]: File sched.c, Line 209 (sched_settime): Request to schedule in the past?!?! Aug 27 09:22:05 NOTICE[1081364]: File sched.c, Line 209 (sched_settime): Request to schedule in the past?!?! Thanks, David Harris
I have the same proble on RH8.0 I use cdr_csv and I do not load aopen. It happens fairly if my system runs for about 24hours. I cannot connect to it with asterisk -r because of broken pipe. I have to kill asterisk process. Serge>From: "David Harris" <davidh@webunited.net> >Reply-To: asterisk-users@lists.digium.com >To: <asterisk-users@lists.digium.com> >Subject: [Asterisk-Users] RE: Asterisk stops responding >Date: Thu, 28 Aug 2003 14:18:00 -0400 > > >Are you using cdr_mysql module ? (storing CDRs in mysql ?) > > >regards > >Martin > >No I am using cdr_csv > >And a "show modules" confirms this. > >David Harris > >_______________________________________________ >Asterisk-Users mailing list >Asterisk-Users@lists.digium.com >http://lists.digium.com/mailman/listinfo/asterisk-users_________________________________________________________________ Add photos to your e-mail with MSN 8. Get 2 months FREE*. http://join.msn.com/?page=features/featuredemail
I have seen similar problems with using Asterisk as a voip UA (though not as bad or predictable as you.) The *8# bug was causing the bulk of my problems. http://bugs.digium.com/bug_view_page.php?bug_id=0000116 David Harris wrote:>Anyone have any thoughts on why versions of asterisk I try (4 so far) >after CVS-07/18/03 always end up locking up on me... which means no sip >clients can register/re-register and if I type "reload" or "stop now" at >the cli it just returns and does nothing. > >I have experienced this same issue on three separate boxes. Two running >RedHat 9 and one running Redhat 8. > >I don't have any digium cards installed. I use SIP only. I use cisco >IP phones and cisco sip voice gateways. I have experienced the issue >while using ztdummy and zaprtc as well as neither. > >The lockup occurs once every 2 hours or so during heavy use, which is >like 50 calls an hour and people checking their voicemail and such. >During off hours like the weekend it usually doesn't even crash at all >even though we do have some call volume. probably like 20 calls all >weekend. > >CLI commands such as "sip show peers", "show channels", "show modules", >still return results. > >Thanks, > >David Harris > >_______________________________________________ >Asterisk-Users mailing list >Asterisk-Users@lists.digium.com >http://lists.digium.com/mailman/listinfo/asterisk-users > >
>I have the same proble on RH8.0 >I use cdr_csv and I do not load aopen. It happens fairly if my systemruns>for about 24hours. I cannot connect to it with asterisk -r because of >broken >pipe. I have to kill asterisk process.This problem is different from mine. I can still reconnect to asterisk with "asterisk -r" and still issue some commands. But I cannot issue either "reload" or "stop now" they return immediately and do nothing. /davidh
>The problem was in cdr_mysql.c since someone send patches that wouldcall>a global reload/unload function. So that's why the reload was locking.>It's fixed in version 1.14 of cdr_mysql.cI am not using cdr_mysql I am using cdr_csv and a "show modules" confirms this. It is more than just "reload" locked, sip clients cannot register or re-register for that matter. /davidh
This is getting to be a big problem. I am hoping it is something
I have setup wrong somewhere...
Various channels just freeze. It always appears to be the agents
phones only. They will come to me and say the phones are down again.
This morning here is what I see. I can not do STOP NOW. Just returns
to
the CLI prompt. I have to kill it. Notice that I try to hangup the
channels and
nothing happens.
Any suggestions?
===========================================================
pbx*CLI> show channels
Channel (Context Extension Pri ) State Appl.
Data
Zap/66-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/54-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/26-1 (macro-enqueue s 105 ) Up Queue
PillNetwork|t||pillnetwork
Zap/25-1 (macro-enqueue s 105 ) Up Queue
PillNetwork|t||pillnetwork
Zap/52-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/65-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/64-1 (local 7100 1 ) Up AgentLogin
(Empty)
7 active channel(s)
pbx*CLI> soft hangup Za
Zap/25-1 Zap/26-1 Zap/52-1 Zap/54-1 Zap/64-1 Zap/65-1 Zap/66-1
pbx*CLI> soft hangup Zap/
Zap/25-1 Zap/26-1 Zap/52-1 Zap/54-1 Zap/64-1 Zap/65-1 Zap/66-1
pbx*CLI> soft hangup Zap/
Zap/25-1 Zap/26-1 Zap/52-1 Zap/54-1 Zap/64-1 Zap/65-1 Zap/66-1
pbx*CLI> soft hangup Zap/66-1
Requested Hangup on channel 'Zap/66-1'
pbx*CLI> soft hangup Zap/65-1
Requested Hangup on channel 'Zap/65-1'
pbx*CLI> soft hangup Zap/64-1
Requested Hangup on channel 'Zap/64-1'
pbx*CLI> soft hangup Zap/54-1
Requested Hangup on channel 'Zap/54-1'
pbx*CLI> soft hangup Zap/52-1
Requested Hangup on channel 'Zap/52-1'
pbx*CLI> soft hangup Zap/25-1
Requested Hangup on channel 'Zap/25-1'
pbx*CLI> soft hangup Zap/26-1
Requested Hangup on channel 'Zap/26-1'
pbx*CLI> show channels
Channel (Context Extension Pri ) State Appl.
Data
Zap/66-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/54-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/26-1 (macro-enqueue s 105 ) Up Queue
PillNetwork|t||pillnetwork
Zap/25-1 (macro-enqueue s 105 ) Up Queue
PillNetwork|t||pillnetwork
Zap/52-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/65-1 (local 7100 1 ) Up AgentLogin
(Empty)
Zap/64-1 (local 7100 1 ) Up AgentLogin
(Empty)
7 active channel(s)
John
1) Is this from current CVS ???
2) does the agent notice by the fact that they can't do a * to hang up the
channel, in fact all dtmf is not recognized
3) if you do a hard hang up the agent line does it stay up
4) Does it only happend when the call is masqueraded to the agent line
5) if the remote hangs up the channel does the agent line come free
If you ans yes to these items, I beleive I have duplicated this in testing
last night
with a config I was testing last night ...
or
is this a system wide deadlock ?? Can you do any other * functions
outside of queues and agents, like dial an extension etc
-----Original Message-----
From: John Congdon <john@z1g.com>
To: asterisk-users@lists.digium.com <asterisk-users@lists.digium.com>
Date: September 3, 2003 5:43 AM
Subject: [Asterisk-Users] Asterisk Stops Responding
>This is getting to be a big problem. I am hoping it is something
>I have setup wrong somewhere...
>
>Various channels just freeze. It always appears to be the agents
>phones only. They will come to me and say the phones are down again.
>
>This morning here is what I see. I can not do STOP NOW. Just returns
>to
>the CLI prompt. I have to kill it. Notice that I try to hangup the
>channels and
>nothing happens.
>
>Any suggestions?
>
>============================================================
I have tried it with a timeout and without...
here the * output for the first side:
-- Starting simple switch on 'Zap/3-1'
-- Executing Dial("Zap/3-1",
"IAX2/useranme:passwd@62.180.50.212/99033283077731") in new stack
-- Called thaeger:hds71@62.180.50.212/99033283077731
-- Call accepted by 62.180.50.212 (format ALAW)
-- Format for call is ALAW
-- Hungup 'IAX2[62.180.50.212:4569]/2'
== No one is available to answer at this time
-- Executing Hangup("Zap/3-1", "") in new stack
== Spawn extension (guersel, 033283077731, 2) exited non-zero on
'Zap/3-1'
-- Hungup 'Zap/3-1'
and here from the other side:
-- Accepting AUTHENTICATED call from 217.81.111.2, requested format = 8,
actual format = 8
-- Executing SetCallerID("IAX2[thaeger@217.81.111.2:4569]/1",
"033283077731") in new stack
-- Executing Dial("IAX2[thaeger@217.81.111.2:4569]/1",
"Zap/g3/033283077731") in new stack
-- Called g3/033283077731
-- Channel 1, span 3 got hangup
-- Hungup 'Zap/63-1'
== No one is available to answer at this time
-- Executing Hangup("IAX2[thaeger@217.81.111.2:4569]/1",
"") in new
stack
== Spawn extension (voipout, 99033283077731, 3) exited non-zero on
'IAX2[thaeger@217.81.111.2:4569]/1'
-- Hungup 'IAX2[thaeger@217.81.111.2:4569]/1'