Hey Alex, Thanks, that does make a lot of sense. I’ve certainly gotten myself into some situations before using the powerful and dangerous toolset that is Kamailio! I will look into the PBX doing this, however it’s a bit tricky, as it’s not aware of the specific requirements of the next hop, or even what the next hop is.
Regardless though, I’m not completely sure if the location of the Content-Length header is the issue here. I’ve manipulated both Content-Type and Content-Length to the bottom of the headers with the following code just to see if this does solve the problem, and it doesn’t appear to: replace_contact_header(“<sip:[email protected]>") cl = KSR::PV.gete("$cl") cT = KSR::PV.gete("$cT") KSR::HDR.remove("Content-Type") if KSR::HDR.is_present('Content-Type') > 0 KSR::HDR.append("Content-Type: #{cT}\r\n") KSR::HDR.remove("Content-Length") if KSR::HDR.is_present('Content-Length') > 0 KSR::HDR.append("Content-Length: #{cl}\r\n") Record-Route: <sip:1.2.3.4;lr;ftag=e4ZXUgGac1e8K;vst=AAAAAAAAAAAAAAAAAAAAAAoUXVFHDQAVd93nfAEGwAAFgwCQA8QY3RlZC5jb20uYXU-;did=923.76f1> ... Call-ID: 814ec57f-e433-83ca-6d8e-023563fb337c CSeq: 17740239 INVITE sip:4.5.6.7:5060;alias=10.100.1.1~5060~1Contact: <sip:[email protected] <sip:[email protected]>> Content-Type: application/sdp Content-Length: 483 Given this is now in line with RFC location of the Content-Length, doesn’t this indicate an issue beyond correct placement of this header? It still seems like the Contact header is still being removed incorrect by the function. FYI I ran a test call with just the KSR::HDR.remove("Contact") call, and still saw the issue: Record-Route: <sip:1.2.3.4;lr;ftag=j8428v06y47jj;vst=AAAAAAAAAAAAAAAAAAAAAAoUXVFHDQGfdDfEGEAEGwAAFgwCQA8QY3RlZC5jb20uYXU-;did=ead.25c1> Call-ID: 0e349984-b83a-1238-6d8e-023563fb337c CSeq: 17740572 INVITE sip:4.5.6.7:5060;alias=10.100.1.1~5060~1Content-Type: application/scp Content-Length: 483 Thanks! Andrew > On 19 Mar 2020, at 2:01 pm, Alex Balashov <[email protected]> wrote: > > Hi Andrew, > > On Thu, Mar 19, 2020 at 01:53:53PM +1100, Andrew White wrote: > >> The substitute and arbitrary position options look useful. In these >> cases, will Kamailio automatically recalculate the Content-Length, or >> do I need to manually rewrite this somehow? > > The Content-Length pertains to the length of an (optional) encapsulated > message body only, and not to the length of the SIP message as such, and > therefore does not apply to any of its headers. > > Encapsulated bodies are used to transport messages of another protocol > inside SIP. In SIP land, the most common encapsulated payload is the SDP > payload, e.g. > > ... > Content-Type: application/sdp > Content-Length: xxx > > [SDP of length xxx goes here] > > If an encapsulated body is not present, Content-Length is still > mandatory but must be set to 0. > >> I did read that in researching before emailing, from similar thread >> you responded to about 5 years ago! In this case, we have a specific >> upstream provider who reject our Contact header in the normal format >> we send, and require it to be in a specific format. As such, I can >> manipulate the header on our PBX or on Kamailio. Given Kamailio is >> aware of the provider we’re sending to and the PBX is not, it seems >> simpler to do on Kamailio. > > The problem with that philosophy is that Contact is fundamentally a > creature of the user agent, not of the proxy, so the proxy shouldn't be > manipulating it. Any time you manipulate it, you run the risk that a > message reliant on that value will be sent in the other direction (the > PBX), and the PBX says, "What?! That's not what I sent!" If that > happens, you rapidly descend into an unmanageable hellscape of stateful > management of this value so as to fool the PBX into thinking that you > didn't change it, and so forth. > > This should really be done on the calling UA if at all possible. Ideally > you could somehow tweak the PBX to send the value in the desired format > to begin with. > > Kamailio provides a lot of tools to shoot onesself in the foot with. It > makes things technically possible that shouldn't necessarily be done > just because they are possible. IMHO. :-) > > -- Alex > > -- > Alex Balashov | Principal | Evariste Systems LLC > > Tel: +1-706-510-6800 / +1-800-250-5920 (toll-free) > Web: http://www.evaristesys.com/, http://www.csrpswitch.com/ > > _______________________________________________ > Kamailio (SER) - Users Mailing List > [email protected] > https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users
_______________________________________________ Kamailio (SER) - Users Mailing List [email protected] https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users
