Predetermined delay in reconnecting to failed WAN

[gravenscroft]

The WAN failover settings are fairly basic and straightforward.

This is correct. Put another way, there are no hidden settings: what you see in the UI is what you can do with the product.

[oj88]

Are you sure that the primary WAN is actually failing? It could just be the server you're using to bounce your ping against that is failing, which is often the case if it is busy or there's traffic congestion going to it.

I had a similar problem before. I was pinging a server that I thought would always be up (DNS) and found out that it would drop pings at random times throughout the day. This causes the WAN to failover and bounce between primary and backup. I fixed it by using a more reliable server IP that has better uptime. No false-positives since.

But to answer you questions, I don't think UT/Arista have those facilities. The WAN failover settings are fairly basic and straightforward. You can only alter the hysteresis for when the failover should occur, not the other way around.

[chlaudan]

Hi,

I have dual wan with failover. I monitor the connection quite aggressively (interval=3, timeout=4, threshold= x3 of 10x). My experience is that when a connection fails and then reconnects within a couple of minutes, it usually goes down again in the next few minutes and keeps on cycling until it stabilizes.
Is there any way I can set it so that if a WAN fails it will not reconnect for a predetermined amount of time i.e. 30 min? Or perhaps, even better, if it can begin retesting the failed connection with a different rule requiring it to prove a consistent connection for say 30 min before restting the WAN connection to "Pass".

Thank you!