[asterisk-dev] cdr_tds is not compiling

Federico Alves sales at minixel.com
Wed Feb 8 20:39:46 MST 2006


I filed bug #0006436 in Mantis. A developer called Corydon76 was assigned
and he can not duplicate the issue, probably because he has to modify the
Makefile and include cdr_tds.so at the top, as Edward mentions it. I have my
development box to demo the problem. Strange enough, an older version of
cdr_tds.so loads just fine in Asterisk 1.2.x, provided freetds 0.61 or 0.62
is installed, but fails with 0.63. That's how I am surviving the upgrade to
1.2.x, but we should fix this problem soon. My question is: is there any
single person on this list who is using cdr_tds?

Message: 4
Date: Wed, 8 Feb 2006 22:17:41 -0000
From: "Edward Eastman" <ed at dm3.co.uk>
Subject: RE: [asterisk-dev] help with cdr_tds not compiling
To: "'Asterisk Developers Mailing List'"
	<asterisk-dev at lists.digium.com>
Message-ID: <20060208221726.3D5EB257ABE at smtp.nildram.co.uk>
Content-Type: text/plain;	charset="us-ascii"

My tdsver.h was installed in /usr/include/freetds by a yum install on centos
4) has which I think explains why I needed to define FREETDS_0_63 (I also
had to add cdr_tds.so to the MODS at the top of the makefile - didn't read
down to the freetds stuff ;)

Given that freetds 0.62 is now over 2 years old is there still a need to
support it?  If we do I guess a change should be made to the makefile to
look in this new location as well as the old one.

Ed

-----Original Message-----
From: asterisk-dev-bounces at lists.digium.com
[mailto:asterisk-dev-bounces at lists.digium.com] On Behalf Of Tilghman Lesher
Sent: 07 February 2006 17:47
To: Asterisk Developers Mailing List
Subject: Re: [asterisk-dev] help with cdr_tds not compiling

On Tuesday 07 February 2006 09:29, Edward Eastman wrote:
> I've been doing some work on getting cdr_tds to use the new custom
> cdr variables and I think I probably came across this problem - I'm
> not sure that this is the way it _should_ be done, but if you add the
> line:
>
> #define FREETDS_0_63
>
> somewhere near the top of cdr_tds.c it should work.  There are
> various ifdef's throughout cdr_tds to make it compatible with both
> newer versions and versions < 0.62, and the logic seems to be a bit
> funny in places.  I'm not sure whether that define should be done
> elsewhere (by freetds headers?) or if it's always been designed to
> work manually.

It's actually in cdr/Makefile and it's attempting to detect the version
number in /usr/include/tdsver.h, specified as TDS_VERSION_NO.
Unfortunately, it's a string, rather than a number, which means we
have to translate it to something else in the Makefile.  Given the
somewhat strange versioning, we attempted to make it work with
later bugfix releases of 0.63.  Note that cdr_tds shouldn't attempt
to build at all, if we aren't able to detect a compatible version.

-- 
Tilghman
_______________________________________________
--Bandwidth and Colocation provided by Easynews.com --

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



------------------------------

Message: 5
Date: Wed, 08 Feb 2006 16:32:49 -0600
From: Kyle Bresin <kbresin at go2call.com>
Subject: [asterisk-dev] Early Media on H.323 problem
To: asterisk-dev at lists.digium.com
Message-ID: <43EA7191.5050001 at go2call.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed


We're having an issue with H.323 on asterisk, passing calls to a Cisco 
5350/5400 Gatekeeper.

The calls are initiated on a PC client, the codec is g723 which asterisk 
passes on the gatekeeper.

The issue we're having is that you never hear any ringing.  And using a 
sniffer, it's confirmed that none of that "Early Media" is getting back 
to the client.

Once the call picks up, audio is fine, no problems.

So we just want ringing to work.

I get the following two errors in the asterisk console:
"Unable to find a codec translation path from g723 to slin"
and
"playtones_alloc: Unable to set 'OH323/XXXXX at XX.XX.XX.XX-2b01a33c' to 
signed linear format"

I've tried the following setups to try to make this go away:
codec_h323 on asterisk 1.2.4
codec_h323 on asterisk 1.0.10
codec_oh323 0.7.3 (against Mimas_patch2 libs) on asterisk 1.2.4
codec_oh323 0.7.3 (against pwlib 1.9.1, openh323 1.17.2.) on asterisk 1.2.4

I get the same error on all of them.  Audio works, no ringing.

Or h323 config is default, we enable only g723.1, set the gatekeeper IP, 
register an extension with it, and set the default context.

The default context in extensions.conf that all calls go through is just 
one dial command:
exten => _99998.,1,Dial(H323/11111#${EXTEN:6},60)

Just enough to strip off the "fake" extension, and pass it onto the 
gatekeeper with a "real" extension.  Again, nothing fancy.

I'm still perplexed why the codec "slin" is even being mentioned.  AFAIK 
nothing involved is using it...

Any help, or advice would be appreciated.

Thanks for reading,

kyle.


------------------------------

Message: 6
Date: Wed, 08 Feb 2006 17:08:17 -0600
From: "Kevin P. Fleming" <kpfleming at digium.com>
Subject: Re: [asterisk-dev] Early Media on H.323 problem
To: Asterisk Developers Mailing List <asterisk-dev at lists.digium.com>
Message-ID: <43EA79E1.8080108 at digium.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Kyle Bresin wrote:

> I'm still perplexed why the codec "slin" is even being mentioned.  AFAIK 
> nothing involved is using it...

Yes it is. The H.323 channel is telling Asterisk that the call is 
'ringing', but for some reason Asterisk does not know how to send 
out-of-band ringing to the originating channel. Since that is the case, 
it has to generate inband ringing, which it can't do because it can't 
transcode into the channel's format.

You need to find out the originating channel won't accept out-of-band 
indications.


------------------------------

Message: 7
Date: Thu, 9 Feb 2006 12:25:18 +1100
From: Edwin Groothuis <edwin at mavetju.org>
Subject: Re: [asterisk-dev] Coverity (was: Possible bug in chan_zap.c)
To: Asterisk Developers Mailing List <asterisk-dev at lists.digium.com>
Message-ID: <20060209012518.GA990 at k7.mavetju>
Content-Type: text/plain; charset=us-ascii

> Please open a bug ticket at http://bugs.digium.com/ so that this can be
> tracked and corrected if it needs to be.  This way, it won't get lost.

Digium should have a chat with the people from Coverity
(http://www.coverity.com/nf_main.html) to get a free trial of a run
of their software on the asterisk codebase
(http://www.coverity.com/free_trial/nf_index.html)

Edwin

-- 
Edwin Groothuis      |            Personal website: http://www.mavetju.org
edwin at mavetju.org    |          Weblog: http://weblog.barnet.com.au/edwin/


------------------------------

Message: 8
Date: Wed, 08 Feb 2006 20:49:40 -0600
From: "Kevin P. Fleming" <kpfleming at digium.com>
Subject: Re: [asterisk-dev] Coverity
To: Asterisk Developers Mailing List <asterisk-dev at lists.digium.com>
Message-ID: <43EAADC4.2050709 at digium.com>
Content-Type: text/plain; charset=ISO-8859-1

Edwin Groothuis wrote:

> Digium should have a chat with the people from Coverity
> (http://www.coverity.com/nf_main.html) to get a free trial of a run
> of their software on the asterisk codebase
> (http://www.coverity.com/free_trial/nf_index.html)

We have already talked to them; I don't know the status of things, but
Coverity is not as 'free' as it seems based on my initial reading of the
situation.


------------------------------

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

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


End of asterisk-dev Digest, Vol 19, Issue 24
********************************************




More information about the asterisk-dev mailing list