Udp timeout!!! 911! Help!

[fireman1720]

Back around the 18th of February 2023, my main SIP Trunk for customer service related calls started acting up.

Apparently if there's no traffic over the SIP trunk, Untangle decides to terminate the UDP session on port 5060 to my trunk provider. Once that happens and I can watch it fall out of the session table, when someone tries to call our 1-800 #, the call fails. I literally had no issue with this for the entire 2022 year, up until last month.

When I called untangle support, they said the UDP timeout couldn't be changed from 30 seconds from the GUI. They said if I was familiar with DEBIAN / Linux, I could try changing the timeout myself. HOW STUPID. Every other firewall support this timeout. Why is Untangle so stupid with this?

Does anyone have any idea how to change this setting? I found the right area (I believe) in Untangle, but all of the current files are blank with no times listed. So I didn't want to change anything just yet without asking.

HELP, PLEASE. Before I burn Untangle and go back to my SonicWall.

[road hazard]

I'm by no means an expert but can you put a band-aid on it and try some sort of bypass rule for that traffic?

[fireman1720]

I'm by no means an expert but can you put a band-aid on it and try some sort of bypass rule for that traffic?

I appreciate the reply. I actually already have bypass rules in place. Still not helping.

[gravenscroft]

They said if I was familiar with DEBIAN / Linux, I could try changing the timeout myself.
Does anyone have any idea how to change this setting?

As Support noted, the change would be made in the Debian underpinning, not in any of the areas specific to NG Firewall itself. There's no setting in NGFW because we just use the Debian default.