I am planning an upgrade for a (3) node 7835-H1 cluster running 7.1(5)SU2 to a (3) node cluster running 8.6(2)a on VMs.
Since these servers only support a bridged upgrade I was wondering if I could do the following to reduce the risk and amount of downtime required for the extended "refresh" upgrade.
1. Install 7.1(5)SU2 on a 7845-H1 lab server
2. Perform a DRS restore on this lab server from the production 7.1(5) publisher
3. Upgrade the lab server to 8.6(2)
4. Perform a DRS backup of 8.6(2)
5. Restore the DRS backup on 8.6(2) Publisher VM (pre-built in an isolated environment using the same IP addresses).
6. Build Subscriber VMs in isolated lab environment using the same IP addresses.
7. Switch routing from the production environment to the pre-built lab environment.
I find this very appealing since our original cluster will still be running untouched in case we have to switch back.
Any concerns with this approach?
As far as I can tell I am basically following the bridged procedure outlined in the 8.6(1) release notes.
The only concern that I see is that the subscribers will not be connected to the lab publisher server when I do the upgrade from 7.1 to 8.6.
Is that going to be an issue?
