AlwaysOn and Availability Groups

With SQL Server 2014, Microsoft continues to push the high-availability (HA) and performance (scale-out) bar higher and higher. Extensive HA options such as AlwaysOn Availability Groups and AlwaysOn Failover Cluster Instances, coupled with a variety of Windows Server family enhancements, provide almost everyone with a chance at achieving the mythical five-nines (that is, 99.999% uptime). We’ll dive into the AlwaysOn new features in this chapter. This capability is taking the database world by storm. It is truly the next generation of HA and scale-out for existing and new database tiers of any kind. Some of the concepts and technical approaches in AlwaysOn and availability groups might seem a bit reminiscent of SQL clustering and database mirroring because they are. Both of these earlier features paved the way for what we now know as AlwaysOn and availability groups.


What’s New in SQL Server AlwaysOn and Availability Groups

Now, with a couple of years under their belt with these features, Microsoft is starting to open up several of the previously tight limitations such as the number of secondaries allowed.

  • You can use AlwaysOn and availability groups with complex data managed through FILESTREAM, even when using Remote Blog Storage and FileTable.
  • Up to 8 secondary replicas can be defined for any one availability group. This used to be 4 secondary replicas max.
  • There can be up to 3 Synchronous Commit replicas.
  • And, client applications can achieve failover across multiple subnets (as many as 64) almost as fast as they can achieve failover within a single subnet.