Questions about the overall project:
- What are your RPO and RTO goals?
- Are there financial penalties if we miss the goals? (Like contracts, refunds to customers, etc)
- Does this app have regularly scheduled maintenance windows, or is it 24/7?
- What’s the ballpark size of the data today? In 3 years?
Questions about the team:
- How many DBAs do we have? How many sysadmins?
- Have they managed Windows clusters before? Multi-subnet clusters?
- Do they have PowerShell experience?
- How do we handle Windows and SQL patches?
- What’s the current on call rotation? Are the skills spread evenly through the rotation, or do SQL cluster calls always require one person to wake up?
Questions about the apps:
- Do the apps require cross database transactions?
- Do all databases have to be at the same exact point in time when failing over to an async (DR) replica, or is it okay if some databases have more data loss than others?
- What apps are going to be querying the secondary?
- How much hardware horsepower do we use on the existing SQL Servers?
- Are users currently happy with performance?
- If the secondary’s data is delayed (minutes, hours), is there a financial risk? (If replication breaks, how urgent is it that we fix it?)
Questions about the hardware and licensing:
- What licensing do we have already? SA?
- What SQL Server version are we aiming for?
- Are we deploying on new hardware/VMs, or trying to reuse existing ones?
- Why not the cloud?
- Why AGs, and why not log shipping, failover clusters, database mirroring, etc.?
