The problem we are having is with processing subscription to reg event at the S-CSCF - we process SUBSCRIBE messages in ims_registrar_scscf module and we are seeing to_tag collisions (though never to_tag/from_tag/call_id collisions) - we process the initial SUBSCRIBE that has no to_tag - we calculate the to_tag with:

tmb.t_get_reply_totag(msg, &ttag);

and store this as part of the subscription.

It seems that this can return a to tag that is the same even if the call_id/from_tag of the initial SUBSCRIBE are different.

I'll investigate further today.

Regards
Richard.

On 24 January 2014 18:18, Klaus Darilion <klaus.mailinglists@pernau.at> wrote:

Am 23.01.2014 15:19, schrieb Jason Penton:

Why are we only searching on to-tag? What if there is a collision on the hash AND the to-tag is the same for 2+ different subscriptions. This is even more likely of happening considering that the hash calculation is based only on the callid and to-tag...

any comments? Am I missing something here?

Maybe it depends on the totag generation. If this code handles incoming SUBSCRIBEs and the totag is generated by Kamailio in a unique way, then it should be safe. If to tags are not generated by Kamailio (and maybe in a bad way), then there may be collisions, but then may be collisions also on callid and from-tag, if the clients generate them in a broken way.

regards
Klaus

_______________________________________________
sr-dev mailing list
sr-dev@lists.sip-router.org
http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-dev