Recently added a 4.5 mbps circuit (3 bonded T1's, waiting for 10X10 fiber build out) to an existing 3 mbps circuit and moved a couple of high traffic vpn sites to the 4.5 interface. About once a week, the UTM reports that the new circuit is down (status shows an error condition) and causes the VPN's to failover to the original 3 mbps circuit. This causes some disruption in the RDP sessions running over those VPN's. Typically the service is down for 5 to 10 minutes, then back up.
The vendor shows the circuits clean, no errors. They did however say that they were seeing periods of up to 98% utilization.
What is my best practice configuration for uplink monitoring to ensure that I don't failover unless the circuit is really down? It would be really rare for 3 T1's to go down at the same time. I have adjusted the uplink monitoring to manual and set ping to 60 seconds with 10 second timeout. Should I also consider limiting bandwidth utilization on the circuit to maybe 90%? Or will that make the problem worse?
This thread was automatically locked due to age.
