I will check. The problem is that this installation is exotic and deviant from my usual order of operations, which creates further confusion.
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/Daniel-Constantin Mierla miconda@gmail.com wrote: On 10/2/12 9:45 PM, Alex Balashov wrote: Perhaps the source of my misunderstanding is something to do with branching.
The actual case I am troubleshooting is a 503 generated via t_reply out of a failure route. The complaint is that this 503 contains a modified From header, exposed to the originator of the call. If you created the transaction after calling msg_apply_changes(), then I am quite sure it is how it worked all the time. Try with an older version and see.
Cheers, Daniel
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote: Those originated locally were never using the original value if you called msg apply changes -- you can try with an older version, if it does that, then it is a bug somewhere :-).
Btw, if you called msg apply changes after creating the transaction, then SL is out of scope and tm local reply is using the initial values, but that means the changes are only for that branch, a failure route will get the initial request content.
Cheers, Daniel
On 10/2/12 9:30 PM, Alex Balashov wrote: The expected behaviour is that all replies, both relayed and endogenously originated (via SL or TM), as well as sequential requests directed back to the caller, will preserve the original headers.
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote: Not sure what is 'work as you expect', provided the details I gave in the previous mail related to local replies.
Otherwise, if you don't track all dialogs, could be an issue if some of the replies are ok and some are not.
I will check the uac module and add a parameter to not use dialog module even if loaded, if such option is not already there.
Cheers, Daniel
On 10/2/12 9:11 PM, Alex Balashov wrote: So, you're saying that if all the dialogs were tracked, everything would work as I expect it to?
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote: I think there was an addition to use dialog module to store initial and new values. I don't know if it can be turned off or is automatically when the dialog module is loaded - just to see if that is the cause.
However, if you do msg_apply_changes(), local replies will use the new headers. I'm pretty sure it happened the same before -- e.g., it was no way to revert back so that sl can use older values of the headers. Perhaps you can revert from config, changing the header to initial value that you store in an avp, apply changes and send the reply.
Received and relayed replies should be converted.
Cheers, Daniel
On 10/2/12 8:57 PM, Alex Balashov wrote: I should be clearer: when I say "internally generated", I don't mean it in the sense of 408 or something else opaquely generated by the proxy. I mean internally generated as in originating from the proxy, e.g. something sent with sl_send_reply() or t_reply().
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Alex Balashov abalashov@evaristesys.com wrote: Hello,
I am using the latest cloned 3.3 branch and performing a uac_replace_from() operation. After the operation, I do msg_apply_changes().
My 'uac' modparams are:
modparam("uac", "rr_from_store_param", "vsfc") modparam("uac", "restore_mode", "auto") modparam("uac", "restore_passwd", "xxxxxxxxx")
However, the modified From URI is getting back to the originator, both in internally generated errors, and in relayed replies.
This didn't used to happen before 3.3, but has been outstanding for quite some time (I didn't look at it much until now). Is there some possibility that there is a >= 3.3 bug? Or, have I simply misconfigured something?
Thanks!
-- Alex
-- Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Fax: +1-404-961-1892 Web: http://www.evaristesys.com/, http://www.alexbalashov.com/
_______________________________________________ SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
_______________________________________________ SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
You can make a simple config, relaying to local host port 9 and in failure route send a t_reply() with your code.
Btw, I just added to dialog module a new parameter that allows disabling usage of dialog module to store the values for From/To headers. As you exposed, it might be a chance that not-tracked dialogs with changes to From/To will not be handled properly - I will write another message to be reviewed.
Anyhow, is good to have an option not to use dialog module. I used it many times in memory-only mode, a restart can cause troubles for sure.
To be easy to backport without a risk for conflicts, I split the changes in two commits, one for source code and the other for docs. If you want to try, just cherry-pick this commit in 3.3:
http://git.sip-router.org/cgi-bin/gitweb.cgi/sip-router/?a=commit;h=0d299fea...
Once the issue is sorted out, I will backport the patch if it proves to fix some wrong behavior.
Cheers, Daniel
On 10/2/12 10:05 PM, Alex Balashov wrote:
I will check. The problem is that this installation is exotic and deviant from my usual order of operations, which creates further confusion.
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote:
On 10/2/12 9:45 PM, Alex Balashov wrote:
Perhaps the source of my misunderstanding is something to do with branching.
The actual case I am troubleshooting is a 503 generated via t_reply out of a failure route. The complaint is that this 503 contains a modified From header, exposed to the originator of the call.
If you created the transaction after calling msg_apply_changes(), then I am quite sure it is how it worked all the time. Try with an older version and see.
Cheers, Daniel
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote: Those originated locally were never using the original value if you called msg apply changes -- you can try with an older version, if it does that, then it is a bug somewhere :-).
Btw, if you called msg apply changes after creating the transaction, then SL is out of scope and tm local reply is using the initial values, but that means the changes are only for that branch, a failure route will get the initial request content.
Cheers, Daniel
On 10/2/12 9:30 PM, Alex Balashov wrote:
The expected behaviour is that all replies, both relayed and endogenously originated (via SL or TM), as well as sequential requests directed back to the caller, will preserve the original headers.
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote: Not sure what is 'work as you expect', provided the details I gave in the previous mail related to local replies.
Otherwise, if you don't track all dialogs, could be an issue if some of the replies are ok and some are not.
I will check the uac module and add a parameter to not use dialog module even if loaded, if such option is not already there.
Cheers, Daniel
On 10/2/12 9:11 PM, Alex Balashov wrote:
So, you're saying that if all the dialogs were tracked, everything would work as I expect it to?
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Daniel-Constantin Mierla miconda@gmail.com wrote: I think there was an addition to use dialog module to store initial and new values. I don't know if it can be turned off or is automatically when the dialog module is loaded - just to see if that is the cause.
However, if you do msg_apply_changes(), local replies will use the new headers. I'm pretty sure it happened the same before -- e.g., it was no way to revert back so that sl can use older values of the headers. Perhaps you can revert from config, changing the header to initial value that you store in an avp, apply changes and send the reply.
Received and relayed replies should be converted.
Cheers, Daniel
On 10/2/12 8:57 PM, Alex Balashov wrote:
I should be clearer: when I say "internally generated", I don't mean it in the sense of 408 or something else opaquely generated by the proxy. I mean internally generated as in originating from the proxy, e.g. something sent with sl_send_reply() or t_reply().
-- Alex
-- Sent from my Samsung mobile, and thus lacking in the refinement one might expect from a proper keyboard.
Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Web: http://www.evaristesys.com/
Alex Balashov abalashov@evaristesys.com wrote: Hello,
I am using the latest cloned 3.3 branch and performing a uac_replace_from() operation. After the operation, I do msg_apply_changes().
My 'uac' modparams are:
modparam("uac", "rr_from_store_param", "vsfc") modparam("uac", "restore_mode", "auto") modparam("uac", "restore_passwd", "xxxxxxxxx")
However, the modified From URI is getting back to the originator, both in internally generated errors, and in relayed replies.
This didn't used to happen before 3.3, but has been outstanding for quite some time (I didn't look at it much until now). Is there some possibility that there is a >= 3.3 bug? Or, have I simply misconfigured something?
Thanks!
-- Alex
-- Alex Balashov - Principal Evariste Systems LLC 235 E Ponce de Leon Ave Suite 106 Decatur, GA 30030 Tel: +1-678-954-0670 Fax: +1-404-961-1892 Web: http://www.evaristesys.com/, http://www.alexbalashov.com/
SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list sr-users@lists.sip-router.org http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
-- Daniel-Constantin Mierla -http://www.asipto.com http://twitter.com/#!/miconda -http://www.linkedin.com/in/miconda Kamailio Advanced Training, Berlin, Nov 5-8, 2012 -http://asipto.com/u/kat Kamailio Advanced Training, Miami, USA, Nov 12-14, 2012 -http://asipto.com/u/katu
-- Daniel-Constantin Mierla -http://www.asipto.com http://twitter.com/#!/miconda -http://www.linkedin.com/in/miconda Kamailio Advanced Training, Berlin, Nov 5-8, 2012 -http://asipto.com/u/kat Kamailio Advanced Training, Miami, USA, Nov 12-14, 2012 -http://asipto.com/u/katu
-- Daniel-Constantin Mierla -http://www.asipto.com http://twitter.com/#!/miconda -http://www.linkedin.com/in/miconda Kamailio Advanced Training, Berlin, Nov 5-8, 2012 -http://asipto.com/u/kat Kamailio Advanced Training, Miami, USA, Nov 12-14, 2012 -http://asipto.com/u/katu
-- Daniel-Constantin Mierla -http://www.asipto.com http://twitter.com/#!/miconda -http://www.linkedin.com/in/miconda Kamailio Advanced Training, Berlin, Nov 5-8, 2012 -http://asipto.com/u/kat Kamailio Advanced Training, Miami, USA, Nov 12-14, 2012 -http://asipto.com/u/katu
-
Alex Balashov -
Daniel-Constantin Mierla