Speedfusion WAN connection priority logic


#1

Today our datacenter peplink dropped the vpn tunnel to 4 sites and it would not start back up. We waited a good 30 minutes while the tunnels tried to re-establish themselves. There was no apparent reason for the outage as the WAN links at the datacenter and the branch sites were all healthy. The branch sites and the datacenter all use different WAN providers and the only thing similar between the branch sites that went down is that they’re all in the southeastern US. The datacenter has two WANs that are used for the VPN but the WANs are configured in 1 and 2 priorities. We do this because setting multiple WANs with priority 1 caused latency issues when we first installed the peplinks and haven’t tried going back to that since prioritizing the WANs has worked for us up till now.

We’re not positive what caused the outage but we’re assuming that some provider in the southeast had an outage that wouldn’t allow traffic to be routed to the prioritized WAN at the datacenter. We eventually got the tunnels to re-establish by changing the WAN priority on the datacenter peplink to our secondary WAN.

I would like to request a feature that a health check is added between the WANs that are used to establish a speedfusion tunnel. In our case the tunnels would never have re-established themselves until the provider corrected what ever occurred on their network. I would like to see the peplink try the secondary priority WAN to establish the tunnel.


#2
  1. Can you share the setting of Remote IP Address / Host Names (Network > SpeedFusion > Select SpeedFusion profile > Remote IP Address / Host Names (Optional)) on data center and remote sites?

  2. Have you checked SpeedFusion WAN status (Status > SpeedFusion) on both data center and remote sites when the problem occurred? WAN1 of both sites showed red color?


#3
  1. I’m sending you a PM with the remote IP addresses.

  2. Yes, I checked this while the problem was occurring, both sites showed green for the WAN connections.

The outage is no longer happening as it was repaired around 4pm on friday but I wanted to make you aware of it so that you can find a work around if this happens to anyone else in the future.


#4

Hi Nick,

I didn’t receive your PM. Anyway, the behavior you described is not expected. SpeedFusion failover is depending on the WAN status and the point to point health check status (this health check is automated, within the tunnel). Please open ticket for us to investigate and download the Diagnostic Report asap. Do let our Tech Support know the actual incident date and time.

Thank you for your help.