Forum Discussion
R_Marc
Dec 08, 2014Nimbostratus
I would question the methodology, personally. Not to say what you are doing wouldn't/couldn't work, but seems overly complex for a blue/green.
I would suggest the following as an alternative:
- Take a backup of each node prior to the upgraded service.
- Turn off sync.
- Update the pools to point to the updated service(s)
- If good, re-enable sync. If bad, force active to standby and then either force a sync from the "good" node to the bad, or apply backup to the "bad" node.
- Re-enable sync.