[Home]

Summary:ASTERISK-05346: Asterisk fails to return stale=true when a nonce is stale
Reporter:edgreenberg (edgreenberg)Labels:
Date Opened:2005-10-21 16:42:06Date Closed:2011-06-07 14:02:41
Priority:MajorRegression?No
Status:Closed/CompleteComponents:Channels/chan_sip/Registration
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) mp108-beta1.txt
( 1) mp108-cvs-head.txt
( 2) mp1xx.txt
Description:In the attached trace, an AudioCodes MP108 gets it's nonces out of sync with Asterisk. In this trace, I can't find where Asterisk sends a 'nonce=stale' back to the Audiocodes box.

****** ADDITIONAL INFORMATION ******

Asterisk 1.2 Beta 1
Audiocodes MP108 w/software ver. 4.40.200.371
Comments:By: Olle Johansson (oej) 2005-10-22 04:21:58

Please try with current CVS head. Thank you.
By: Olle Johansson (oej) 2005-10-22 04:23:55

And upload a debug file where you have debug=4 and verbose=4 and sip debug turned on so we see the comments from the auth part of chan_sip. Thank you.
By: Morten Isaksen (misaksen) 2005-10-22 14:36:46

I think this is a duplacte of bug 0004329.
By: edgreenberg (edgreenberg) 2005-10-22 19:04:10

1. 4329 is about ForkCDR. Would misaken please clarify what bug this duplicates, so I can read the original?

2. I will have to set up a test machine with CVS head on it. I'll try to do this Monday afternoon.
By: Morten Isaksen (misaksen) 2005-10-23 02:50:13

Sorry. http://bugs.digium.com/view.php?id=5329
By: edgreenberg (edgreenberg) 2005-10-23 10:58:28

Well, CVS HEAD _seems_ to work, in initial testing. I realize that this should mean, close the bug, but I'd like to understand what patch I can install from after beta1 that will resolve the issue in beta1.  Would one of you bug-fixers be willing to engage with me offline to help me discover this?

First thing I'm going to do is apply 5329, though I have my doubts that it is the problem, since the "stale=true" string never appears in the output from Asterisk.

Many thanks,
</edg>
By: Morten Isaksen (misaksen) 2005-10-23 11:05:21

If you look at the output from the Audiocodes box you can that it thinks the nonce is

nonce="14ed711b" "

(notice the double quote). This happens because of the extra space.

I had the same problem some time ago. Check http://bugs.digium.com/view.php?id=5378
By: edgreenberg (edgreenberg) 2005-10-23 11:16:38

Applying the 5329 patch resolves the issue. Many thanks.

I will also communicate this to an Audiocodes SE that I work with.
</edg>
By: Olle Johansson (oej) 2005-10-23 11:23:10

Already fixed in cvs. Thank you for quick feedback!