Hello again Stefan,
Just in case anything goes wrong during the update, we’re thinking backing up the databases (including config, content, and service)and snapshot each farm server (excluding SQL server) before the update.
So if halfway through the binary installation something goes wrong, we will restore the servers from the snapshot.
And if the Config Wizard encounters error that we can’t fix, we rollback by restoring the databases, and also restore the servers from their snapshot.
Is this approach OK, or are we completely nuts thinking this is the proper way of rolling back?
Thank you!