SQL Server 2005: I Said Be Cautious, Not Petrified
In a recent article, I stated that people should be cautious in upgrading to SQL Server 2005. Some irate readers have taken that to mean that I think there is nothing worthwhile in the new release.
I can understand why, based on the title and the initial subtitle (which was subsequently changed), a cursory glance could have given that impression.
In any case, I thought I should make myself as clear as possible. SQL Server 2005 has a great many features that would recommend its use.
As a former DBA, I am most excited about the new data-mirroring feature which is similar to Oracle Data Guards physical standby feature, but seemingly better at failback than Data Guard.
It is an improvement over basic log shipping because it streams the active log changes. This could support both high availability or disaster recovery scenarios.
Other HA features I like include online index builds and early restore access, which is another availability feature useful when database recovery is required.
Other things like database snapshots or database views offer some interesting flexibility opportunities such as the offloading of reporting and analysis services.
Suffice it to say I believe there are many excellent features in this release that DBAs will love, especially those that have supported other database platforms that already had similar functionality.
Take data partitioning for example, which has finally been introduced for the first time. This is a key scalability enhancement and manageability enhancement for databases with very large amounts of data.
Microsoft has taken an interesting approach to partitioning by utilizing partitioning functions which administrators can create and associate with a table or index.
This means it is more flexible than simple range or hash partitioning. The security model seems much improved also, including a more granular permission model.
So before any of you SQL Server fans out there get all upset, take a breath. My position is not that SQL Server 2005 is not a significant releasejust that from a business standpoint there is no need to rush to it.
Hey, its new, so its bound to have some issues as all new software does.
I am certainly not alone in this view. Check out Lisa Vaas recent article, which featured some survey results from SQL Server users and their likely plans to make the upgrade. It noted that only 13 percent of the responders plan to upgrade immediately.
Eighty percent stated they need more time and 46 percent stated they would wait at least a year before they would trust it with their mission-critical systems.
What does all of this mean? Simply that adoption cycles for new versions remain about the same even when it takes the vendor six years to come out with another version.
If I wasnt clear in the previous article, let me state that the long release cycle, while undoubtedly unintended, was brilliant on Microsofts part. What can you say? Sometimes you just get lucky.
People appreciate stability, and by George, we had that with SQL Server 2000. Lets all hope that this new version and its associated product, Visual Studio, prove to be solid, relatively stable pieces of software.
In any case, if it were me and my IT processes were doing fine as they are, Id just assume we let that 13 percent figure out where all the problems are.
As a general once said, "Men, Ill be right behind you all the way!"
Charles Garry is an independent industry analyst based in Simsbury, Conn. He is a former vice president with META Groups Technology Research Services. He can be reached at email@example.com.
Check out eWEEK.coms for the latest database news, reviews and analysis.