Summary:	ASTERISK-19173: All blind transfers failing on 1.8.9.0-rc1
Reporter:	Ross Beer (rossbeer)	Labels:
Date Opened:	2012-01-09 05:41:28.000-0600	Date Closed:	2012-01-24 10:06:36.000-0600
Priority:	Blocker	Regression?
Status:	Closed/Complete	Components:	Channels/chan_sip/Transfers
Versions:	1.8.9.0	Frequency of Occurrence	Constant
Related Issues:	must be completed before resolving ASTERISK-19040 Asterisk 1.8.9.0 Blockers must be completed before resolving ASTERISK-19041 Asterisk 10.1.0 Blockers causes ASTERISK-21394 [patch] - Fundamental changes to CDR within single asterisk family (1.8) during externally initiated blind transfers with an h extension present is related to ASTERISK-18811 Dialplan is not processed after AGI script when dst channel hangup first
Environment:	Cent OS	Attachments:	( 0) ASTERISK-19173.patch
Description:	All blind transfers failing since 1.8.9.0 and previous versions are unaffected. As soon as a blind transfer is initiated the channel is hung up.
Comments:	By: David Woolley (davidw) 2012-01-09 05:51:44.683-0600 It is not entirely clear whether or not you are using SIP native transfers, as against features.conf ones. For SIP problems, you need to provide both "sip set debug on" type output and "sip set history on" output, with the associated verbose and debug level logging. By: Ross Beer (rossbeer) 2012-01-09 06:32:49.993-0600 Transfers are being used via SIP native transfers on SIP phones. I will provide the additional information shortly. By: Matt Jordan (mjordan) 2012-01-09 07:41:40.612-0600 If you can, go ahead and get a pcap of this as well. By: Matt Jordan (mjordan) 2012-01-12 13:01:08.316-0600 Ross - I've attempted to reproduce this with pjsua and been unable to do so.  To move this forward we're definitely going to need a pcap and, preferably, a DEBUG log illustrating the failure. By: Kristijan Vrban (vrban) 2012-01-17 07:58:28.021-0600 It's  changeset 347595 that introduced this error. But, the question that there is much more important to clarify here: I thought that this kind of these regression of basic functions were caught by automated tests? Or not? And if not, would it really make sense to test such things automatically. Really useful. By: Paul Belanger (pabelanger) 2012-01-17 09:12:59.362-0600 What your last comment really necessary?  Do you feel better now that you have raged? As for the testsuite not catching it, I'm not sure why.  Feel free to research it and report back.  Other wise, have some patience and try to be civil. By: Mark Michelson (mmichelson) 2012-01-20 16:39:18.727-0600 I'm uploading ASTERISK-19173.patch to the issue. This fixed it for me. I've also created a review for the change here: https://reviewboard.asterisk.org/r/1685/ Please let us know if this fixes the issue for you. EDIT: The patch was made against the 1.8 tip, not version 1.8.9.0-rc1. If the patch does not apply properly, let me know and I'll roll a new one for you. By: Kristijan Vrban (vrban) 2012-01-23 08:39:16.683-0600 I can confirm that ASTERISK-19173.patch fix the issue in my setup By: Luke H (luckman212) 2012-01-24 09:06:17.504-0600 I can confirm this patch fixed the issue for me as well.  Great!