I've got an in-house application that we are currently piloting. Typically it takes my company 9-18 months to move from pilot to 100% implementation. Because of this delay, our operations group prefers to start projects with "starter" hardware before scaling out to the required hardware for full implementation.
I'm looking to make my life easier, and I'd like to know if the following migration is feasible:
On our existing [starter] server, create an Availability Group with a listener.
Route all SQL traffic through that AG listener (as opposed to directly to the sql instance).
Purchase 2 servers, SAN, etc. Build cluster, and setup the cluster as a secondary for the existing AG.
Then I would just need to make the cluster the primary and ultimately remove the original server as a secondary. (how would I do this?)
Additionally I would add 2 more [new] R/O secondaries for reporting.
Can anyone comment on this as a plan?