Hi,

You can achieve this.

Say if you are failing over to standby 2 at SCN A,

Make sure Standby 1 is not ahead of SCN A, to make sure this stop the recovery on standby 1 before failover.

Perform failover.

Remove/defer log_archive_dest_2 and  log_archive_dest_3 on primary to defer the remote archiving to standby 1 and 2.

Setup parameters correctly on standby 2 after failover, start managed recovery on standby 1.

Standby 1 will pick the redo from new primary and will become standby for new primary (standby 2 )

Regards

Anudeep

if you are using dgmgrl, life is easier.you can say just "switchover to "db unique name of standby"

If you use sqlplus to manage the standby, this would be case

primary
stdby1 - in managed recovery
stdby2 - - in managed recovery

Initiate the switchover from primary

alter database commit to switchover to physical standby with session shutdown;
shu immediate
startup mount

Both standby willreceive the end of redo from  primary Mrp will apply and terminate by itself

Go to the desired standby, let's say

stdby2

alter database commit to switchover to primary;
shu immediate
startup

Now stdby2 has to ship redo to primary and stdby1 mount the stdby1 too and start the managed recovery in both primary and stdby1

For example,

I assume the below db_unique_names for your dg setup. Primary directly sends redo to both standby databases

primary
stdby1
stdby2

Let's say you want to make stdby2 as a new primary

You should have started the managed recovery in the below manner

** on stdby1 **

alter database recover managed standby database through all switchover
disconnect;

** stdby2 **

alter database recover managed standby database disconnect;

Now start the switchover from primary

** on pirmary ***

alter database commit to switchover to physicl standby with session shutdown;
shu immediate
startup mount

** on stdby2 **

alter database commit to switchover to primary;
shu immediate
startup

** on pirmary ***

alter database recover managed standby database disconnect;

stdby1 does not require any manual intervention. It will apply the new archived redo logs coming from new primary

Regards,

nd