Clustering - WSFC vs AGs

  • Hi,

    More and more I'm having discussions with customers who want to move away from FCI to AGs. The main reasoning is that AGs offer "faster, almost instant" failover times. I will admit I've never delved in to this difference, so I am asking here.

    What are the under-the-hood pros and cons of each; there are some obvious ones as far as storage is concerned. Basically, what option would be best-used for what requirements?

    Thanks

  • Yes, the additional storage is the main drawback, although that stops disk from being the only single point of failure.  You also need to be careful to match up the configuration on both (or all sides), for example by creating the same SQL logins with the same SIDs on each server.  Like you say, almost instant failover is a big advantage - you can do OS and SQL Server patching with next to no downtime.  And, since you don't have to have every database in your availability group, you have more control over what gets failed over.

    John

  • Thanks!

    Yeah that's what I mean, there's no option that has all the pros and none of the cons!

    Is the instant failover thing really that much of an issue with the advent of Hyper-Convergence and Cluster shared volumes (I know that's 2014 up only)? This is basically what this post is about - real-world timings!

    My personal preference is for WSFC because of the single storage resource and as you say "config match up". The only pro I see from AG is being able to pick and choose what DBs you want to protect.

  • I assume you are using Enterprise Edition, not Standard Edition; otherwise, you can only put one database in one basic AG.

    If merge replication is used(as my employer does), move publisher to AG will face other issues.

Viewing 4 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic. Login to reply