5 Oct
2010
5 Oct
'10
3:46 p.m.
Hi!
The TM docs give this example (#59):
at http://kamailio.org/docs/modules/stable/modules/tm.html#t_reset_fr
failure_route[0]{
if (!t_branch_timeout()){
if (t_any_timeout()){
log("one branch did timeout\n");
sl_send_reply("408", "Timeout");
}
}
}
If I use sl_send_reply in a failure route (like in the example) I get
0(19813) : <core> [cfg.y:3333]: parse error in config file
/etc/kamailio/kamailio.cfg, line 719, column 49: Command cannot be used in the block
Now, is the documentation or the sl module in error?
/O :-)
5 Oct
5 Oct
5:42 p.m.
Olle E. Johansson writes:
If I use sl_send_reply in a failure route (like in the
example) I get
0(19813) : <core> [cfg.y:3333]: parse error in config file
/etc/kamailio/kamailio.cfg, line 719, column 49: Command cannot be used in the block
Now, is the documentation or the sl module in error?
perhaps you should try send_reply() function instead of sl_send_reply.
-- juha
5:53 p.m.
5 okt 2010 kl. 16.42 skrev Juha Heinanen:
Olle E. Johansson writes:
So the docs are in error...
I used t_reply() which worked...
Thanks, Juha.
/O
If I use sl_send_reply in a failure route (like
in the example) I get
0(19813) : <core> [cfg.y:3333]: parse error in config file
/etc/kamailio/kamailio.cfg, line 719, column 49: Command cannot be used in the block
Now, is the documentation or the sl module in error?
perhaps you should try send_reply() function instead of sl_send_reply. 
6:43 p.m.
2010/10/5 Olle E. Johansson <oej(a)edvina.net>et>:
failure_route[0]{
if (!t_branch_timeout()){
if (t_any_timeout()){
log("one branch did timeout\n");
sl_send_reply("408", "Timeout");
}
}
}
If I use sl_send_reply in a failure route (like in the example) I get
0(19813) : <core> [cfg.y:3333]: parse error in config file
/etc/kamailio/kamailio.cfg, line 719, column 49: Command cannot be used in the block
Now, is the documentation or the sl module in error?
It must be an error in the documentation. When entering in
failure_route this implies that the transaction does exist, so
sl_send_reply() is not valid.
--
Iñaki Baz Castillo
<ibc(a)aliax.net>
6:47 p.m.
5 okt 2010 kl. 17.43 skrev Iñaki Baz Castillo:
2010/10/5 Olle E. Johansson <oej(a)edvina.net>et>:
Hmm. I am not sure you are right here. You enter the failure_route when a
transaction fails, but it still exists. You can retrigger it with new branches. Or has it
ceased to exist so you simply create a new transaction?
We need to add this to the docs too.
/O
failure_route[0]{
if (!t_branch_timeout()){
if (t_any_timeout()){
log("one branch did timeout\n");
sl_send_reply("408", "Timeout");
}
}
}
If I use sl_send_reply in a failure route (like in the example) I get
0(19813) : <core> [cfg.y:3333]: parse error in config file
/etc/kamailio/kamailio.cfg, line 719, column 49: Command cannot be used in the block
Now, is the documentation or the sl module in error?
It must be an error in the documentation. When entering in
failure_route this implies that the transaction does exist, so
sl_send_reply() is not valid.
6:59 p.m.
2010/10/5 Olle E. Johansson <oej(a)edvina.net>et>:
I mean the incoming transaction (that created upon receipt of the
INVITE). t_relay is executed (the transaction is created). When
entering in failure_route that transaction obviously exists (we
haven't sent a final reply to the caller yet) so using sl_send_reply
is invalid (it can be used just in case the transaction hasn't be
created).
It must be an
error in the documentation. When entering in
failure_route this implies that the transaction does exist, so
sl_send_reply() is not valid.
Hmm. I am not sure you are right here. You enter the failure_route when a
transaction fails, but it still exists. You can retrigger it with new branches.
Yes, but it has nothing to do with the incoming transaction which is unique.
Or has it ceased to exist so you simply create a new
transaction?
The incoming transaction exists until a final response is sent by
kamailio to the client. Outgoing transactions (branches) exist until
kamailio receives a final reply for each one.
--
Iñaki Baz Castillo
<ibc(a)aliax.net>
5166
days inactive
5166
days old
7 comments
3 participants
participants (3)
-
Iñaki Baz Castillo -
Juha Heinanen -
Olle E. Johansson