With all of the variables now in play, upgrading requires much more thought.
For example, we have a client with 20 users on 6.3.3, that doubling the cores (2 to 4) with 12G of RAM found v8.0.2 to be unacceptable when they tried it in a test environment. For them, we are still testing to find the right resource allocation that meets their expectations.
Another client went from 6.3.3 to 7.1.2 (50-70 users), and after going to Full SQL (3 cores) and 32G of RAM, they are stable, but replacing that server with a new server with more cores and more RAM that we are currently configuring. More details after we transition them.
I could go on, but the bottom line is that practices need to think it through, and perhaps test (or have someone like us provide them a test) before they upgrade.
We are working with AC to refine some of this into a methodology, and I'll be putting out a series of videos going through this in detail since this is now the #1 subject we are getting questions about.
It is much deeper subject, but it needs to be pointed out that not all VM technology is the same when it comes to efficiency or performance.
