Summary

The crux of high availability is in laying a fundamentally sound foundation that you can count on when failures occur. Then, when failures do occur, determining how much data loss you can tolerate, how much downtime is possible, and what this downtime is costing you are all part of understanding your high availability picture.

The future seems to be improving greatly in all of the foundation levels' support for high availability. This includes

• Cheaper and more reliable hardware components that are highly swappable

• The advent of virtual server capabilities (with Windows Virtual Server 2005) to insulate software failures from affecting hardware

• Enhancements that Microsoft is making to SQL Server 2005 that address availability square in the face

The critical enhancements to the cornerstone availability capabilities of SQL clustering will help this fault-tolerant architecture grow more reliable for years to come. The big bonuses come with the new features of database mirroring as another fault-tolerant solution at the database level along with the new database snapshots feature to make data more available to more users more quickly than the older methods of log shipping.

And last but not least, the enhancements to the everyday operations of restores, restarts, partitioning, indexing, and the new snapshot isolation level all directly affect availability in one or more ways. We will still have to wait a bit before these new technologies are available and considered stable, but our high availability future is getting brighter and more easily achievable for even the smallest application of the smallest company.