SIP NAT Helper - still causes "problems", or was that fixed ages ago?

Discussion in 'Tomato Firmware' started by BikeHelmet, Oct 22, 2013.

  1. BikeHelmet

    BikeHelmet Addicted to LI Member

    Hi all,

    I've been optimizing my VOIP setup recently - combing the internet for articles and forum posts with info on tweaking my PAP2T and SPA112 to get them working better. So far it's made a big difference to call quality.

    One thing that I keep coming across is that Tomato's SIP NAT Helper "causes problems" - but the posts and articles never go into more detail than that. Only that it "causes problems" and to turn it off - is this old advice that is still being parroted even though things have changed? I've got all the helpers turned on and everything has always worked properly. (Except for latency issues and stuff, which I fixed by tweaking the devices and my QOS setup.)

  2. koitsu

    koitsu Network Guru Member

    Until someone provides a full in-detail write-up of what "causes problems" means -- that means, show me the packet contents and show me what this "SIP helper" actually is doing and what issue it's causing -- I'm inclined to believe it's not causing problems at all.
  3. BikeHelmet

    BikeHelmet Addicted to LI Member

  4. koitsu

    koitsu Network Guru Member

    That isn't how I interpret it at all. In the first URL, the individual with the last post is saying that the SIP helper capability in netfilter is not available in Linux 2.4 but is available in Linux 2.6.

    So, on "K26" TomatoUSB firmwares, you will see this capability; it's a GUI checkbox under Advanced / Conntrack/Netfilter.

    None of this gives me any detailed information about what the supposed "problem" is with it. :) So until someone can provide a highly detailed technical write-up with actual packet captures and point out what the issue is with the SIP helper, I'm inclined to dismiss such claims as nonsense.

    I'm quite familiar, for example, with what the FTP helper does -- netfilter has to actually modify the payload of the packets after they're defragmented, because the FTP protocol in active (not passive) mode actually includes the IP address of the client (i.e. 192.168.x.x) within the packet itself, hence why it needs to be rewritten (with the WAN address). IRC DCC operates the same way.
  5. BikeHelmet

    BikeHelmet Addicted to LI Member

    My ISP keeps giving me back the same IP, so I doubt I'll be able to test whether this is an issue. Regardless, doesn't seem to be affecting me. I'll assume it's been "fixed" until someone proves otherwise.
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice