[SR-Users] Re: Using http_async_query - transaction seems "reused" for subsequent SIP INVITE requests received ? (Kamailio 5.7.3)

On Jan 12, 2024, at 9:26 AM, Chaigneau, Nicolas &lt;nicolas.chaigneau(a)capgemini.com&gt; wrote: Thanks for the clarification. So "branches" is not what I need. So I should do something like that: append_to_reply("Contact: <...>\r\n"); sl_send_reply("302", "Moved Temporarily"); exit; (I do have an "exit" immediately after all calls to "sl_send_reply") I tried with a minimalist request_route, as follows: request_route { if (is_method("INVITE")) { append_to_reply("Contact: <sip:+1234;myparam=test@$si:$sp>\r\n"); xlog("L_INFO", "Calling sl_send_reply then exit\n"); sl_send_reply("302", "Moved Temporarily"); exit; } } But I still have this WARNING message logged by Kamailio when handling a SIP INVITE: WARNING: <core> [core/dset.c:690]: print_dset(): no r-uri or branches Regards, Nicolas. -----Message d'origine----- De : Alex Balashov via sr-users &lt;sr-users(a)lists.kamailio.org&gt; Envoyé : vendredi 12 janvier 2024 13:44 À : Kamailio - Users Mailing List Cc : Alex Balashov Objet : [SR-Users] Re: Using http_async_query - transaction seems "reused" for subsequent SIP INVITE requests received ? (Kamailio 5.7.3) ******This mail has been sent from an external source. Do not reply to it, or open any links/attachments unless you are sure of the sender's identity.****** Hi Nicolas, Yes, I think that unfortunately this is the outcome of some confusion. Apart from the word "append", there is nothing in common between the concepts of append_branch() and append_to_reply(). I'm not sure why you are getting the messages you are just sending redirects, but the prime suspicion is that the execution of your configuration contains additional steps beyond just sending stateless replies, e.g. that you call sl_send_reply(), but do not call 'exit' afterward, and so the config execution progresses to a step where some kind of relay is attempted. Conceptually, sending redirects is as simple as: request_route { ... # Maybe some sanity-checking or request boilerplate here. if(method == "INVITE") { # some DB ops or whatever, yielding $var(val). append_to_reply("Contact: <sip:$var(val)>\r\n"); sl_send_reply("302", "Moved Temporarily"); exit; } # Anything else that might occur in this config should not # occur if an INVITE was received--note the 'exit' step above. } It may not be quite as simple as that, but hopefully this gives the right idea. -- Alex > On 12 Jan 2024, at 03:16, Chaigneau, Nicolas &lt;nicolas.chaigneau(a)capgemini.com&gt; wrote: > > Hello Alex, > > > The confusion is probably on my part. > > Reading this: > https://kamailio.org/docs/modules/devel/modules/sl.html#sl.f.sl_send_ > r > eply > > 3.1. sl_send_reply(code, reason) > For the current request, a reply is sent back having the given code and text reason. The reply is sent stateless, totally independent of the Transaction module and with no retransmission for the INVITE's replies. > > If the code is in the range 300-399 (redirect reply), the current destination set is appended to the reply as Contact headers. The destination set contains the request URI (R-URI), if it is modified compared to the received one, plus the branches added to the request (e.g., after an append_branch() or lookup("location")). If the R-URI was changed but it is not desired to be part of the destination set, it can be reverted using the function revert_uri(). > > Custom headers to the reply can be added using append_to_reply() function from textops module. > > > > I thought that I needed to use append_branch before calling sl_send_reply to control the Contact headers in the reply. > > I tried to use "append_to_reply" instead to add the Contact headers, like this: > append_to_reply("Contact: <...>\r\n"); > > This works, but then I get WARNING messages in the logs: > WARNING: <core> [core/dset.c:690]: print_dset(): no r-uri or branches > > Which is also why I was confused... > You're telling me I should not create branches... but if I don't, I get these messages. > > Could you please clarify ? > > > Thanks a lot. > > Regards, > Nicolas. > > > -----Message d'origine----- > De : Alex Balashov via sr-users &lt;sr-users(a)lists.kamailio.org&gt; Envoyé : > jeudi 11 janvier 2024 18:57 À : Kamailio (SER) - Users Mailing List > Cc > : Alex Balashov Objet : [SR-Users] Re: Using http_async_query - > transaction seems "reused" for subsequent SIP INVITE requests > received ? (Kamailio 5.7.3) > > ******This mail has been sent from an external source. Do not reply > to it, or open any links/attachments unless you are sure of the > sender's > identity.****** > > Hi, > > First off, a bit confused as to why you are appending a branch and then sending a final reply? Adding a branch only makes sense if you plan to fork the request to an additional destination, instead of responding to the sender with a final dispositive (>= 3xx) reply. > > -- Alex > >>> On 11 Jan 2024, at 12:16, Chaigneau, Nicolas via sr-users &lt;sr-users(a)lists.kamailio.org&gt; wrote: >> >> Hello, >> So far I was using Kamailio in "stateless" mode to handle SIP INVITE requests and reply with 302. >> I am now trying to use module http_async_client module, but I'm experiencing unexpected behavior with "branches". >> I'm using function http_async_query as described in the example: >> >> https://www.kamailio.org/docs/modules/devel/modules/http_async_client. >> html#http_async_client.f.http_async_query >> When the transaction is resumed, I'm building and sending the reply, using "append_branch" and "sl_send_reply": >> https://kamailio.org/docs/modules/devel/modules/sl.html#sl.f.sl_send >> _ >> r >> eply >> For example: >> append_branch("..."); >> sl_send_reply("302", "Moved Temporarily"); This >> works, however when I'm sending new client SIP INVITE requests to Kamailio, it seems it will always reuse the previous transaction. >> The new branches are appended to the branches of the first transaction. >> I end up with errors "ERROR: <core> [core/dset.c:424]: append_branch(): max nr of branches exceeded" when the limit (12) is exceeded. >> I do not understand why this happens. This is a new SIP INVITE message, it should not be linked to the previous transaction ? >> I tried a few things: >> - remove the transaction using "t_release();" >> - configure: modparam("tm", "wt_timer", 0) This did not help... >> How can I solve this ? >> Thanks for your help. >> Regards, >> Nicolas. >>