UDP Timeout Setting / SIP helper service

It seems that when Sip helper service is activated, connection tracking keep UDP streams connections during one hour instead of the normal UDP timeout setting.

This cause troubles when using VoIP even without NAT. (specially with dynamic IP SIP clients or multiple Internet Gateways).

The workaround is to manually remove unreplied connections inside the connection tracking list. Not a very nice solution…


Is it possible to correct this, or at least give a setting for UDP timeout when using SIP helper ?


The actual implementation of the SIP Helper certainly give lots of problems to users.

If this is not corrected, then it would be better to disable it by default, and use RTP static rules in the firewall instead of this dangerous toy.

I think I looked at it before and it detected the registration interval negotiated with the SIP connection. IE: If I changed it on my phones, it would change in con tracking. I cant remember for sure. I have all my phones set to 60 second keepalives and they always work.

The problem is that even if it detect the registration interval, you can have a dynamic IP address changing during this interval.

Then the connection is lost for hours.

I guess I dont see what you are getting at. One of my techs has a remote phone at his home office. Before we switched to running the VOIP over the VPN tunnels, if his PPPOE connection dropped and reconnected, his phone would simply reconnect to the PBX.