Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-13 Thread Steve Davies

On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:

I'm interested to understand why I many messages like:

WARNING[21314] chan_zap.c: Ring requested on channel 0/1 already in use
on span 1.  Hanging up owner

How can a channel be already in use??? That means the channel is
busy...if it is so then it is all right...but maybe that shouldn't be a
warning but a notice or something else...should it?



What this means is that a call has arrived from your Telco on a Zap
channel. Asterisk THINKS that this channel is already in use - In my
case, asterisk is wrong. Sadly, asterisk hangs up the new call (so
no-one can call you anymore!) and does not have an option to clean up
the call data to show the channel as free. Outbound calls are fine,
but will never be placed on this hung channel.

IMHO, 99% of the time, if the telco says a channel is free, they know best!!!

As per my other email, I see this problem on one site. Sadly I have
not been able to identify where it comes from (although I suspect
broken SIP clients upsetting Asterisk)

What versions of zaptel/asterisk do you use? What hardware, and what
phone devices? Perhaps people with this problem have something in
common.

To start the ball rolling:
 Software: Zaptel 1.0.9, Asterisk 1.0.9, BRIstuff-0.2.0, wanpipe-2.3.2
 PRI interface: Sangoma A101U (UK E1)
 Phones on sites with NO problems: snom, elmeg, Aastra, Linksys/Sipura
 Phones on problem site: Hitachi WIP3000, Zyxel F1000 (?)

Regards,
Steve
___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-13 Thread Giorgio Incantalupo

Hi Steve,

I agree with you..telco knows better!
If telco sends a ring on channel X  and asterisk has already used it, 
couldn't asterisk shift that call on another channel Y or it is 
obliged to answer on channel X?
In other words, if asterisk get a ring on channel 3 and channel 3 is in 
use, Asterisk should use channel 6 to answer the call and so connect 
channel 3 to 6. Isn't it? Or I have not correctly understood how telco 
and Asterisk speak between them?


My PBX configuration is (may change due to upgrades):

Debian Sarge 3.1 r0a
asterisk-install: v6
Asterisk: 1.2.9.1
Zaptel: 1.2.6
Libpri: 1.2.3
BRI: install-misdn-mqueue 0.3.1-rc23



Giorgio Incantalupo



Steve Davies wrote:

On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:

I'm interested to understand why I many messages like:

WARNING[21314] chan_zap.c: Ring requested on channel 0/1 already in use
on span 1.  Hanging up owner

How can a channel be already in use??? That means the channel is
busy...if it is so then it is all right...but maybe that shouldn't be a
warning but a notice or something else...should it?



What this means is that a call has arrived from your Telco on a Zap
channel. Asterisk THINKS that this channel is already in use - In my
case, asterisk is wrong. Sadly, asterisk hangs up the new call (so
no-one can call you anymore!) and does not have an option to clean up
the call data to show the channel as free. Outbound calls are fine,
but will never be placed on this hung channel.

IMHO, 99% of the time, if the telco says a channel is free, they know 
best!!!


As per my other email, I see this problem on one site. Sadly I have
not been able to identify where it comes from (although I suspect
broken SIP clients upsetting Asterisk)

What versions of zaptel/asterisk do you use? What hardware, and what
phone devices? Perhaps people with this problem have something in
common.

To start the ball rolling:
 Software: Zaptel 1.0.9, Asterisk 1.0.9, BRIstuff-0.2.0, wanpipe-2.3.2
 PRI interface: Sangoma A101U (UK E1)
 Phones on sites with NO problems: snom, elmeg, Aastra, Linksys/Sipura
 Phones on problem site: Hitachi WIP3000, Zyxel F1000 (?)

Regards,
Steve
___
--Bandwidth and Colocation provided by Easynews.com --

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



___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-13 Thread Steve Davies

On 9/13/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:

Hi Steve,

I agree with you..telco knows better!

If telco sends a ring on channel X  and asterisk has already used it,
couldn't asterisk shift that call on another channel Y or it is
obliged to answer on channel X?


The telco is in charge and is not asking, it is telling asterisk
that there is a call on channel X. If asterisk thinks this channel is
busy, then it is wrong because it must agree that a channel is busy
(SETUP/SETUP_ACK) with the telco first.

Probably, a now finished call was made on that channel, and some old
data in asterisk has not been cleaned-up when the call went away.


In other words, if asterisk get a ring on channel 3 and channel 3 is in
use, Asterisk should use channel 6 to answer the call and so connect
channel 3 to 6. Isn't it? Or I have not correctly understood how telco
and Asterisk speak between them?


BRI/PRI channels are like point to point data pipes, they cannot be
cross connected in that way. If the telco knows that channel 3 is busy
and channel 6 is available, it will just use channel 6 in the first
place.

(I stand ready to be told I am wrong here)


My PBX configuration is (may change due to upgrades):

Debian Sarge 3.1 r0a
asterisk-install: v6
Asterisk: 1.2.9.1
Zaptel: 1.2.6
Libpri: 1.2.3
BRI: install-misdn-mqueue 0.3.1-rc23


No obvious common factor then...

Steve
___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-12 Thread Giorgio Incantalupo

Problema solved!

Just put resetinterval=never inside zapata.conf


Giorgio Incantalupo


Giorgio Incantalupo wrote:

Hi,
I installed an Asterisk box with a sangoma A102 PRI card. Sometimes 
Asterisk drops calls...there is nothing inside logs but these warnings:


Sep 11 15:00:18 WARNING[3503] chan_zap.c: Ring requested on channel 
0/6 already in use on span 1.  Hanging up owner.
Sep 11 15:00:22 WARNING[3503] chan_zap.c: Got restart ack on channel 
0/6 span 1 with owner
Sep 11 15:00:30 WARNING[3503] chan_zap.c: Ring requested on channel 
0/3 already in use on span 1.  Hanging up owner.
Sep 11 15:29:38 WARNING[3497] chan_sip.c: Maximum retries exceeded on 
transmission [EMAIL PROTECTED]

.168.3.175 for seqno 2 (Critical Response)
Sep 11 15:30:04 WARNING[3497] chan_sip.c: Maximum retries exceeded on 
transmission [EMAIL PROTECTED]

2.168.3.175 for seqno 2 (Critical Response)
Sep 11 15:30:24 WARNING[3497] chan_sip.c: Maximum retries exceeded on 
transmission [EMAIL PROTECTED]

2.168.3.175 for seqno 2 (Critical Response)
Sep 11 15:31:34 WARNING[3503] chan_zap.c: Ring requested on channel 
0/6 already in use on span 1.  Hanging up owner.
Sep 11 16:00:08 WARNING[3503] chan_zap.c: Ring requested on channel 
0/6 already in use on span 1.  Hanging up owner.
Sep 11 16:00:26 WARNING[3503] chan_zap.c: Got restart ack on channel 
0/6 span 1 with owner
Sep 11 16:03:13 WARNING[3503] chan_zap.c: Ring requested on channel 
0/6 already in use on span 1.  Hanging up owner.

Sep 11 16:03:15 WARNING[15530] app_dial.c: Unable to forward voice
Sep 11 16:07:09 WARNING[3503] chan_zap.c: Ring requested on channel 
0/6 already in use on span 1.  Hanging up owner.

Sep 11 16:13:22 WARNING[15925] app_dial.c: Unable to forward voice
Sep 11 16:15:37 WARNING[3503] chan_zap.c: Ring requested on channel 
0/6 already in use on span 1.  Hanging up owner.

Sep 11 16:21:26 WARNING[3504] chan_zap.c: Call specified, but not found?
Sep 11 16:23:21 WARNING[16267] app_dial.c: Unable to forward

Anyone ever got these messages? What do they mean? How can I fix them?

TIA

Giorgio Incantalupo

___
--Bandwidth and Colocation provided by Easynews.com --

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



___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-12 Thread Steve Davies

For the curious, can anyone tell me how this flag fixes the issue? - I
have seen the error before, but always assumed it was related to hung
channels.

Thanks,
Steve

On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:

Problema solved!

Just put resetinterval=never inside zapata.conf


Giorgio Incantalupo


Giorgio Incantalupo wrote:
 Hi,
 I installed an Asterisk box with a sangoma A102 PRI card. Sometimes
 Asterisk drops calls...there is nothing inside logs but these warnings:

 Sep 11 15:00:18 WARNING[3503] chan_zap.c: Ring requested on channel
 0/6 already in use on span 1.  Hanging up owner.
 Sep 11 15:00:22 WARNING[3503] chan_zap.c: Got restart ack on channel
 0/6 span 1 with owner
 Sep 11 15:00:30 WARNING[3503] chan_zap.c: Ring requested on channel
 0/3 already in use on span 1.  Hanging up owner.
 Sep 11 15:29:38 WARNING[3497] chan_sip.c: Maximum retries exceeded on
 transmission [EMAIL PROTECTED]
 .168.3.175 for seqno 2 (Critical Response)
 Sep 11 15:30:04 WARNING[3497] chan_sip.c: Maximum retries exceeded on
 transmission [EMAIL PROTECTED]
 2.168.3.175 for seqno 2 (Critical Response)
 Sep 11 15:30:24 WARNING[3497] chan_sip.c: Maximum retries exceeded on
 transmission [EMAIL PROTECTED]
 2.168.3.175 for seqno 2 (Critical Response)
 Sep 11 15:31:34 WARNING[3503] chan_zap.c: Ring requested on channel
 0/6 already in use on span 1.  Hanging up owner.
 Sep 11 16:00:08 WARNING[3503] chan_zap.c: Ring requested on channel
 0/6 already in use on span 1.  Hanging up owner.
 Sep 11 16:00:26 WARNING[3503] chan_zap.c: Got restart ack on channel
 0/6 span 1 with owner
 Sep 11 16:03:13 WARNING[3503] chan_zap.c: Ring requested on channel
 0/6 already in use on span 1.  Hanging up owner.
 Sep 11 16:03:15 WARNING[15530] app_dial.c: Unable to forward voice
 Sep 11 16:07:09 WARNING[3503] chan_zap.c: Ring requested on channel
 0/6 already in use on span 1.  Hanging up owner.
 Sep 11 16:13:22 WARNING[15925] app_dial.c: Unable to forward voice
 Sep 11 16:15:37 WARNING[3503] chan_zap.c: Ring requested on channel
 0/6 already in use on span 1.  Hanging up owner.
 Sep 11 16:21:26 WARNING[3504] chan_zap.c: Call specified, but not found?
 Sep 11 16:23:21 WARNING[16267] app_dial.c: Unable to forward

 Anyone ever got these messages? What do they mean? How can I fix them?


___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-12 Thread Rich Adamson

Steve Davies wrote:

For the curious, can anyone tell me how this flag fixes the issue? - I
have seen the error before, but always assumed it was related to hung
channels.

Thanks,
Steve

On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:

Problema solved!

Just put resetinterval=never inside zapata.conf


Giorgio Incantalupo


If memory serves correctly, I believe the parameter was added a couple 
of years ago as a means / workaround for hung channels. At the time, 
there was not any overwhelming evidence as why a channel would 
occasionally hang. Some of the possibilities included unusual 
interaction from the opposite end of the T1/E1, anomalies in the 
dialplan, etc.


Now that a substantial amount of work / changes have been made relative 
to PRI's and other internal asterisk code, there appears to be less of a 
need to reset.


A reasonable approach might be to apply the parameter and pay close 
attention to channels that might be in some strange state. If none are 
observed, then leave it.


___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-12 Thread Steve Davies

On 9/12/06, Rich Adamson [EMAIL PROTECTED] wrote:

Steve Davies wrote:
 For the curious, can anyone tell me how this flag fixes the issue? - I
 have seen the error before, but always assumed it was related to hung
 channels.

 Thanks,
 Steve

 On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:
 Problema solved!

 Just put resetinterval=never inside zapata.conf


 Giorgio Incantalupo

If memory serves correctly, I believe the parameter was added a couple
of years ago as a means / workaround for hung channels. At the time,
there was not any overwhelming evidence as why a channel would
occasionally hang. Some of the possibilities included unusual
interaction from the opposite end of the T1/E1, anomalies in the
dialplan, etc.

Now that a substantial amount of work / changes have been made relative
to PRI's and other internal asterisk code, there appears to be less of a
need to reset.

A reasonable approach might be to apply the parameter and pay close
attention to channels that might be in some strange state. If none are
observed, then leave it.


Thanks for that. I have a customer who is using Asterisk 1.0.x, and I
am tempted to backport this fix from the 1.2.x code where it was
introduced.

Cheers,
Steve
___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-12 Thread Rich Adamson

Steve Davies wrote:

On 9/12/06, Rich Adamson [EMAIL PROTECTED] wrote:

Steve Davies wrote:
 For the curious, can anyone tell me how this flag fixes the issue? - I
 have seen the error before, but always assumed it was related to hung
 channels.

 Thanks,
 Steve

 On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:
 Problema solved!

 Just put resetinterval=never inside zapata.conf


 Giorgio Incantalupo

If memory serves correctly, I believe the parameter was added a couple
of years ago as a means / workaround for hung channels. At the time,
there was not any overwhelming evidence as why a channel would
occasionally hang. Some of the possibilities included unusual
interaction from the opposite end of the T1/E1, anomalies in the
dialplan, etc.

Now that a substantial amount of work / changes have been made relative
to PRI's and other internal asterisk code, there appears to be less of a
need to reset.

A reasonable approach might be to apply the parameter and pay close
attention to channels that might be in some strange state. If none are
observed, then leave it.


Thanks for that. I have a customer who is using Asterisk 1.0.x, and I
am tempted to backport this fix from the 1.2.x code where it was
introduced.


From a personal perspective, I think I'd hold off on the back port and 
devote that time towards testing the soon to be released version (now in 
Trunk).


If you've watched the number and type of changes that have gone into SVN 
Trunk in the last couple of months, it appears as though a significant 
number of possible memory leaks, sip code, infrastructure code, PRI code 
changes, etc, have been applied that would be beneficial for all 
production systems. There also appears to be a fair amount of work that 
will be needed to upgrade dialplan syntax (etc) for the new release.


Best guess is that once the Trunk code gets past the beta testing phase, 
it will likely be the asterisk code of choice for most/all production 
systems.


Consider the above is only my $0.02 worth. ;)

___
--Bandwidth and Colocation provided by Easynews.com --

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


Re: [asterisk-users] PRI: sometimes Asterisk drop calls

2006-09-12 Thread Giorgio Incantalupo

Hi,
thanks to all
I solved the calls dropped problem, it was resetinterval parameter in 
zapata.now asterisk does not drop calls anymore.

I do not get the message:

WARNING[3503] chan_zap.c: Got restart ack on channel 0/6 span 1 with owner

anymore...but I get all the others.
I'm interested to understand why I many messages like:

WARNING[21314] chan_zap.c: Ring requested on channel 0/1 already in use 
on span 1.  Hanging up owner


How can a channel be already in use??? That means the channel is 
busy...if it is so then it is all right...but maybe that shouldn't be a 
warning but a notice or something else...should it?



TIA


Giorgio Incantalupo



Rich Adamson wrote:

Steve Davies wrote:

On 9/12/06, Rich Adamson [EMAIL PROTECTED] wrote:

Steve Davies wrote:
 For the curious, can anyone tell me how this flag fixes the issue? 
- I

 have seen the error before, but always assumed it was related to hung
 channels.

 Thanks,
 Steve

 On 9/12/06, Giorgio Incantalupo [EMAIL PROTECTED] wrote:
 Problema solved!

 Just put resetinterval=never inside zapata.conf


 Giorgio Incantalupo

If memory serves correctly, I believe the parameter was added a couple
of years ago as a means / workaround for hung channels. At the time,
there was not any overwhelming evidence as why a channel would
occasionally hang. Some of the possibilities included unusual
interaction from the opposite end of the T1/E1, anomalies in the
dialplan, etc.

Now that a substantial amount of work / changes have been made relative
to PRI's and other internal asterisk code, there appears to be less 
of a

need to reset.

A reasonable approach might be to apply the parameter and pay close
attention to channels that might be in some strange state. If none are
observed, then leave it.


Thanks for that. I have a customer who is using Asterisk 1.0.x, and I
am tempted to backport this fix from the 1.2.x code where it was
introduced.


From a personal perspective, I think I'd hold off on the back port and 
devote that time towards testing the soon to be released version (now 
in Trunk).


If you've watched the number and type of changes that have gone into 
SVN Trunk in the last couple of months, it appears as though a 
significant number of possible memory leaks, sip code, infrastructure 
code, PRI code changes, etc, have been applied that would be 
beneficial for all production systems. There also appears to be a fair 
amount of work that will be needed to upgrade dialplan syntax (etc) 
for the new release.


Best guess is that once the Trunk code gets past the beta testing 
phase, it will likely be the asterisk code of choice for most/all 
production systems.


Consider the above is only my $0.02 worth. ;)

___
--Bandwidth and Colocation provided by Easynews.com --

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



___
--Bandwidth and Colocation provided by Easynews.com --

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