Yitzhak Bar Geva
2007-Oct-22 23:35 UTC
[asterisk-users] NAT traversal packet loss measurement
How can one measure the effect of NAT traversal packet loss? We currently have no solution for NAT traversal for our SIP clients. There is no doubt that packets are getting lost. What is not clear is how much damage this does. On the face of it, everything seems fine. Could this be so? Perhaps we're suffering a degradation in quality or our call setup times could be improved. How can we measure this? What's the simplest method of preventing packet loss due to NAT traversal in a SIP environment? Thanks, Yitzhak -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20071023/13da9c41/attachment.htm
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Yitzhak Bar Geva wrote:> How can one measure the effect of NAT traversal packet loss? > We currently have no solution for NAT traversal for our SIP clients. There > is no doubt that packets are getting lost. What is not clear is how much > damage this does. On the face of it, everything seems fine. Could this be > so? Perhaps we're suffering a degradation in quality or our call setup times > could be improved. How can we measure this? > What's the simplest method of preventing packet loss due to NAT traversal in > a SIP environment?NAT is unlikely to cause a percentage of packets to get lost. Normally you'd have one way audio if NAT was causing a problem (i.e. 100% packet loss). The only other situation in which it might happen is where the NAT router decides to close a port mapping (thereby blocking incoming calls to the customer's device). But if you're looking for packet loss there are a number of other things to check first. I wouldn't do VoIP across the WAN without at least some packet shaping but hey. - -- Kind Regards, Matt Riddell Director _______________________________________________ http://www.venturevoip.com (Great new VoIP end to end solution) http://www.venturevoip.com/news.php (Daily Asterisk News - html) http://feeds.venturevoip.com/AsteriskNews (Daily Asterisk News - rss) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFHHUC+DQNt8rg0Kp4RAv0uAJ9Q41eQ+7RuqzFvgtxEhQOIU0QFggCaAlkD GMVdY/n58wHsciuHihZCCHY=6L87 -----END PGP SIGNATURE-----
Yitzhak Bar Geva wrote:> How can one measure the effect of NAT traversal packet loss? > We currently have no solution for NAT traversal for our SIP clients.We've recently completed a setup (see other threads) with a couple of SIP clients behind NAT in their respective home-offices. Took a couple of attempts, but after consulting the list, we have a working setup.> What's the simplest method of preventing packet loss due to NAT > traversal in a SIP environment?I doubt very much if any loss you're seeing is due to NAT traversal. /Per Jessen, Z?rich -- http://www.spamchek.com/ - your spam is our business.