Customers has App server (MS Server 2012) and Storage server (MS SQL 2012), with several hundred clients accessing server app via VPN.
Now they want to have a high-availability and also load balancing if possible, by setting up another same instance (App & Storage servers, 2012). But due some reasons the preferred location is another DC, and actual (measured) bandwidth availible to us is ~400mbps between these 2 sites (existing site and preferred one). As usual, the budget is also constrained :)
By digging, I found
A) "Peer-to-peer transactional replication" option, which as I understand, gives near-real time replication and also load balancing (but requires MS SQL Enterprise edition)
B)" AlwaysOn failover cluster instances" (WSFC) option (for 2 nodes, MS SQL Standard Edition is OK). It is HA setup, but will not provide load-balancing.
The question is:
Is assumptions on A) & B) correct (HA, NLB)?
Is 400mbps bandwidth between 2 physical instances enough? It is assumed that daily SQL transactions/load is 5-10GB (SQL server network traffic).
I have very limited understanding on MS SQL and DB servers, so maybe other more optimal solutions? (but have to stick to MS SQL 2012)
0 Answers