So, its late. SQL Server 2005, Yukon, is not going to ship until mid-2005. So what? Who cares? Users are actually downright gleeful that theyre going to get a major new release thats more stable than a two-legged stool.
For me, a forum entry by SJ, requiring registration to view, on SQLServerCentral.com really nails the issue. Not only is it crucial to have a stable and secure SQL Server environment just for the sake of not going nuts with business continuity/high availability/data protection/what have you, its important for Microsoft to come out with a product that doesnt let the Oracle-inclined score any points. SJ writes: “I want and NEED a stable environment. With all the enhancements coming (database mirroring, .Net integration, the awesome workbench … ) I am a bit worried about the stability of the product. Also, SECURITY should be a major concern! In my environment, I am constantly competing with another division that is set on Oracle. So far, I have been able to fend them off with several factors and features and being able to integrate with them. They have seen how stable our setup is, and that is helping. I dont want to lose any ground because the next version isnt stable and secure.”
So if it comes in mid-2005, thats cool. If it comes late 2005, what the heck. As long as its secure, and as long as its stable.
The lateness issues
But there are other issues at stake regarding the calendar. For one thing, theres the money spent for years of upgrade assurance from the Software Assurance program. All these years have gone by without any upgrades, even incremental. True, as pointed out to me by Tom Rizzo, director of product management for SQL Server, Microsoft did deliver some nifty stuff outside of upgrades. There was the Web Services tool kit for SQL Server 2000 that the Redmondites put out in 2002. That provided Web services support and rich XML support. There was the new version of SQL Server CE, which was nice for customers who wanted a rich mobile database. (At least, Microsoft said it was nice for those customers. Was it? Let me know if it wasnt.) And then there were Reporting Services, an undeniably cool and unarguably right-priced (read, free) business-intelligence tool.
And its also true that the Software Assurance program delivers other things besides upgrades that made DBAs lives easier. For the money, you get support, training, deployment tools, employee discounts on software and extended warrantees.
But make no mistake about it, users are still hungry for the new features SQL Server 2005 promises to deliver. Both Microsoft and SQLServerCentral.com have been running surveys to find out which features users are most eager to get their hands on. From what Im gleaning, data mirroring is by far one of the top picks. Theres a hodgepodge of other features users are excited about, including the dedicated administrator connection, new and improved T-SQL and an integrated tool set—if it has good monitoring and management tools a la SQLProbe, according to a posting by Aethyr Dragon.
The cutting of enterprise
-class features”> But one thing thats worthy of note is the lack of interest in those features that should have made SQL Server 2005 a real competitor in the enterprise database space, including clustering within the DBMS (shared disk or shared nothing), hash partitioning, SQL Server 2005 as the unifying storage engine within Microsoft applications, and file system support within the DBMS. Those are things that Microsoft had promised from the get-go, when it first started talking about the update, and theyre features that would give SQL Server supporters some useful fodder in the inter-IT department SQL Server/Oracle wars. As Gartners Betsy Burton pointed out to me, those are also the features that would enable SQL Server stalwarts to convince upper management that the database is ready to take on more, bigger, more mission-critical jobs, thus enabling SQL Server to reach its tentacles deeper into the enterprise than it has to date. Unfortunately, their chances for making it into the final product are just about nil at this point.
Another calendar issue is, of course, continued support for SQL Server 2000 and 7.0, which is slated to come to a crashing halt on Dec. 31, 2005. If SQL Server 2005 ships in mid-2005, that gives customers all of—yikes!—six months to upgrade. True, Microsoft has conscientiously dealt with extending support in the past, but as the days tick past, were still waiting to hear about it this time. Customers are getting—well, not panicky, exactly, but you could say theyre a bit nervous. Heres another post from SQLServerCentral.com that epitomizes the angst:
“Support a product for only six months after the latest version comes out? WTF! This is very NOT COOL, and Microsoft better clarify soon. For many shops, upgrading within a year will be an issue, much less six months. And how bout software that is SQL2K-dependent? Theyll have to develop an upgrade version before their customers can upgrade dependent databases, and theres no way that can happen in six months. Also, what about businesses that cant afford to upgrade immediately? They will be left high and dry as far as support is concerned. Microsoft better watch out; the time frame for upgrading to SQL2K5 will be the same time frame that mySQL will be taking off its training wheels. If theyre not extremely cool about this, they risk losing customers.”
MySQL as a threat to one of the database Big Boys? Well, this isnt the first time the possibility has been mentioned. So, Microsoft, WTH! About that extended support? Were all ears.
Let me know which SQL Server 2005 feature is going to make a difference to you. Also, Id love to hear from anybody who was messed up by the delay—particularly ISVs that were working toward a 2004 release. Write me at firstname.lastname@example.org.
eWEEK.com Database Center Editor Lisa Vaas has written about enterprise applications since 1997.
Be sure to add our eWEEK.com database news feed to your RSS newsreader or My Yahoo page: