|Summary:||ASTERISK-03867: chan_iax2 not RFC compliant :p|
|Date Opened:||2005-04-05 09:49:56||Date Closed:||2011-06-07 14:10:29|
|Description:||On page 15, it says a full frame is sent when the 16 bit timestamp wraps around.|
We are trying to hunt down some other issues, while looking at ethereal traces we found out that in the case of:
iaxclient -> * -> * -> pstn, the iaxclient is sending the full frame and asterisk will ACK it but asterisk itself is not sending a full frame when going over the 65535ms.
I talked about this with stevek, he thinks asterisk should send it. (i also interprete the RFC in the same way.)
****** ADDITIONAL INFORMATION ******
Im not going to make a patch for this, we first want to find the other things in iax2 (trunking related) that are causing me problems.
Feel free to comment, patch or close out.
I did not test this with -HEAD.
|Comments:||By: Mark Spencer (markster) 2005-04-05 14:49:21|
This is an area of the spec that should probably have some discussion. In a sense, this behavior should probably be refined.
By: Kevin P. Fleming (kpfleming) 2005-04-29 13:27:02
Is this still an issue?
By: Michael Jerris (mikej) 2005-05-23 23:26:47
zoa, there were many jb inspired timestamp refining patches commited in the last few weeks. Is this resolved now or do we need to address it still?
By: Russell Bryant (russell) 2005-05-24 04:53:40
Well, he said that he only tested this with 1.0. I do not have the capability to test right now, so someone needs to lab this up and see if it is still an issue with CVS HEAD. If so, then we can proceed from there to decide waht the appropriate behavior should be. Once we clear this up in CVS HEAD, I will make the necessary changes to 1.0.
By: Tilghman Lesher (tilghman) 2005-06-26 13:37:51
Closed due to lack of response by poster. Reopen if this is still an issue.