Hello, Thanks for the hint to use cherry-pick,
After applying these two patches the issue seems to be fixed. Please find out the attachment with kamailio logs.
Thank you for your help.
Regards José
2016-06-21 10:41 GMT+01:00 Daniel-Constantin Mierla miconda@gmail.com:
Hello,
you should cherry-pick them with git into the branch 4.4.
Cheers, Daniel
On 21/06/16 11:04, José Seabra wrote:
Hi Daniel, I have applied your last two patchs on tm module but i can't compile the module because kemi framwork.
I'm using kamailio version 4.4.
Compile Errors:
t_fwd.c:61:24: error: ../../kemi.h: No such file or directory t_fwd.c: In function ‘prepare_new_uac’: t_fwd.c:157: error: ‘sr_kemi_eng_t’ undeclared (first use in this function) t_fwd.c:157: error: (Each undeclared identifier is reported only once t_fwd.c:157: error: for each function it appears in.) t_fwd.c:157: error: ‘keng’ undeclared (first use in this function) t_fwd.c:346: warning: implicit declaration of function ‘sr_kemi_eng_get’ t_fwd.c:348: warning: implicit declaration of function ‘sr_kemi_act_ctx_get’ t_fwd.c:348: warning: assignment makes pointer from integer without a cast t_fwd.c:350: warning: implicit declaration of function ‘sr_kemi_act_ctx_set’ t_fwd.c:352: warning: implicit declaration of function ‘sr_kemi_cbname_lookup_idx’ t_fwd.c: In function ‘add_blind_uac’: t_fwd.c:715: error: ‘TM_UAC_FLAG_BLIND’ undeclared (first use in this function) make: *** [t_fwd.o] Error 1
Regards José
2016-06-21 9:38 GMT+01:00 José Seabra joseseabra4@gmail.com:
Hello Daniel,
I confirm that the phone receives 200 OK to the CANCEL and 487 to the INVITE.
I will backport the tm module to the last two patchs, then, after make all tests I will report back to you.
Regards José
2016-06-21 5:33 GMT+01:00 Daniel-Constantin Mierla < miconda@gmail.com miconda@gmail.com>:
Can you try with master branch or backport the last two patches from tm module? I pushed two commits that should catch and handle better this case.
Cheers, Daniel
On 20/06/16 18:14, Daniel-Constantin Mierla wrote:
Hello,
it seems it tries to generate an outgoing cancel for the suspended branch. I will check the code, likely there has to be added condition for this cases.
Is the 487 reply for invite sent back? Also, the 200ok for cancel?
Cheers, Daniel
On 20/06/16 16:38, José Seabra wrote:
Hello,
I'm attaching more logs to this email regarding to the issue on SIP CANCEL to an INVITE that is suspended.
If do you think that i should open an issue on git regarding to this let me know.
Thank you for your help.
Best Regards José
2016-06-15 14:42 GMT+01:00 José Seabra < joseseabra4@gmail.com joseseabra4@gmail.com>:
Hi Daniel,
But when Kamailio receives a CANCEL prints the following error messages:
2016-06-15 14:39:10.354 ERROR: tm [t_msgbuilder.c:287]: build_local_reparse(): ERROR: build_local_reparse: INVITE is missing 2016-06-15 14:39:10.354 ERROR: tm [t_msgbuilder.c:494]: build_local_reparse(): ERROR: build_local_reparse: cannot build CANCEL request 2016-06-15 14:39:10.354 ERROR: tm [t_cancel.c:310]: cancel_branch(): ERROR: attempt to build a CANCEL failed 2016-06-15 14:39:10.354 ERROR: tm [t_fwd.c:1389]: e2e_cancel(): ERROR: cancel error
I'm handling the CANCEL in the script by the following way:
if (is_method("CANCEL")) { if (t_check_trans()) { route(RELAY); } else { sl_send_reply("481", "Call leg/transaction does not exist"); } exit(); ... ... ...
Thank you for your support.
Regards José
2016-06-15 12:15 GMT+01:00 Daniel-Constantin Mierla < miconda@gmail.commiconda@gmail.com>:
Hello,
On 14/06/16 16:33, José Seabra wrote:
Hi Olle and Daniel, Thank you for your replies, After receive your msg I looked again to my script and i found the problem.
I didn't configure the correct failure_route block and the failure route configured didn't print any msg on the logs, so I thought that it wasn't entering on failure route.
Sorry for my mistake.
Anyway, How should i handle the CANCEL sip msg to an INVITE that is suspended? (still related with this implementation)
Just handling it as done in the default configuration file is ok -- the suspended transaction will be canceled.
Cheers, Daniel
-- Daniel-Constantin Mierlahttp://www.asipto.com - http://www.kamailio.orghttp://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
-- Cumprimentos José Seabra
-- Cumprimentos José Seabra
-- Daniel-Constantin Mierlahttp://www.asipto.com - http://www.kamailio.orghttp://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda