12 Nov
2007
12 Nov
'07
6:03 p.m.
Hello all,
is it possible to completely drop a reply from failure_route? I know it is
possible to change the reply with t_reply, but I need the possibility to send
nothing back to the UAC.
--
Greetings,
Alex Hermann
12 Nov
12 Nov
6:19 p.m.
Hi Alex,
No this in not possible. In statefull mode, the server has not send a
reply to the client.
regards,
bogdan
Alex Hermann wrote:
Hello all,
is it possible to completely drop a reply from failure_route? I know it is
possible to change the reply with t_reply, but I need the possibility to send
nothing back to the UAC.
13 Nov
13 Nov
10:44 a.m.
Isn't it possible to drop provisional responses?
http://www.openser.org/dokuwiki/doku.php/core-cookbook:devel#drop
klaus
Bogdan-Andrei Iancu schrieb:
Hi Alex,
No this in not possible. In statefull mode, the server has not send a
reply to the client.
regards,
bogdan
Alex Hermann wrote:
Hello all,
is it possible to completely drop a reply from failure_route? I know it is
possible to change the reply with t_reply, but I need the possibility to send
nothing back to the UAC.
_______________________________________________
Users mailing list
Users(a)lists.openser.org
http://lists.openser.org/cgi-bin/mailman/listinfo/users
11:41 a.m.
Hi Klaus,
yes, it is possible to drop provisional replies. But I think Alex was
referring to final negative reply as he mentioned failure route (you do
not get any provisional reply there).
Regards,
Bogdan
Klaus Darilion wrote:
Isn't it possible to drop provisional responses?
http://www.openser.org/dokuwiki/doku.php/core-cookbook:devel#drop
klaus
Bogdan-Andrei Iancu schrieb:
Hi Alex,
No this in not possible. In statefull mode, the server has not send a
reply to the client.
regards,
bogdan
Alex Hermann wrote:
Hello all,
is it possible to completely drop a reply from failure_route? I know
it is possible to change the reply with t_reply, but I need the
possibility to send nothing back to the UAC.
_______________________________________________
Users mailing list
Users(a)lists.openser.org
http://lists.openser.org/cgi-bin/mailman/listinfo/users
12:41 p.m.
On Tuesday 13 November 2007 09:41, Bogdan-Andrei Iancu wrote:
Hi Klaus,
yes, it is possible to drop provisional replies. But I think Alex was
referring to final negative reply as he mentioned failure route (you do
not get any provisional reply there).
Right.
I pass REGISTER requests to an upstream registrar, but when it is unavailable
I use the 'save()' function from failure_route to create a valid 200 OK
response to the UAC. In this scenario the UAC also receives the locally
created 408 or upstream response. I want to drop this (negative) response.
Luckily the 200 OK is sent earlier than the 408, but there is no gauarantee is
will arrive earlier. It is also not known how every UAC will react when it
receives both replies.
--
Met vriendelijke groet,
Alex Hermann
Speakup
T: 088-SPEAKUP (088-7732587)
F: 088-7732588
7:06 p.m.
Alex Hermann schrieb:
On Tuesday 13 November 2007 09:41, Bogdan-Andrei Iancu
wrote:
IMO it should be possible to also drop final responses (yes I know it is
bad as the transactions stay in memory but sometimes it is needed).
As a workaround - in failure route loop the request to the proxy again
and then save it (of course src_ip is then wring. Thus for NAT traversal
you have to store the received paramters in a proprietary header which
will be read when receiving the REGISTER again).
regards
klaus
Hi Klaus,
yes, it is possible to drop provisional replies. But I think Alex was
referring to final negative reply as he mentioned failure route (you do
not get any provisional reply there).
Right.
I pass REGISTER requests to an upstream registrar, but when it is unavailable
I use the 'save()' function from failure_route to create a valid 200 OK
response to the UAC. In this scenario the UAC also receives the locally
created 408 or upstream response. I want to drop this (negative) response.
Luckily the 200 OK is sent earlier than the 408, but there is no gauarantee is
will arrive earlier. It is also not known how every UAC will react when it
receives both replies.
6206
days inactive
6207
days old
5 comments
3 participants
participants (3)
-
Alex Hermann -
Bogdan-Andrei Iancu -
Klaus Darilion