Manual failover

we may need a few more details. in your traffic group configuration do you have "Auto Failback" enabled?

N

I assume they're on a pair, if they stop talking to each other, they'll both think the peer went down and the Standby will go Active. Once the communication is restored, one of them will go into Standby.

What exactly is the issue? Would you want A to be the preferred Active unit rather than B? What BIG-IP version is being used?

The problem is that during the failover we have lost sessions and the workload is processed in the secondary data center. It doesn´t make sense my secondary data center to cause some kind of unavailability on the primary data center.

That's right, I'd like A to remain active and failover not occur, but from what I understand, I'll have to enable "Always Failback to First Device if it's Available"

13.0.0 Build 2.0.1671

As you said, when the communication is restored, one of them must be in the standby mode, but it should not be the A, because it is with preferred device order.

that was about to be my next suggestion, to have a look at the preferred device order...

I just checked the Auto Failback help description on v13:

"Specifies whether the traffic group fails back to the initial device specified in Failover Order. When checked, causes the traffic group to fail back to the initial device specified in Failover Order whenever that device is as available or more available than another device in the group. Note that the initial device becomes active even when other devices in the group are more available. When cleared, causes the traffic group to remain active on the current device until that device becomes unavailable".

So it sounds like either you should try to actually enable Auto Failback, or as you mentioned tick the "always Failback to First Device if it is Available" option

I´ll try this option.

Thanks a lot!