With SQLRAP 2.5 out the door a couple of weeks ago, I’ve been free to turn to another project, SIPA, an internal effort to automate storage and retrieval of our group’s diverse intellectual property catalog. I’m acting as both the database architect and one of the database implementers on the project.
This is my first significant from-scratch development effort under SQL Server 2008, and I’m incredibly impressed:
Every time I turn around, it seems as though I find another arrow that the SQL Server dev team has added to my quiver. When one combines the newer tools in synergistic ways, one gets code which is compact and powerful, yet easy to understand.
Coding in SQL Server 2008 is a lot of fun. As I get deeper into SIPA, I’ll share more of what i learn.
this copyrighted material was originally posted at http://blogs.technet.com/wardpond.
the author and his employer are pleased to provide this content for you at that site, and via rss, free of charge and without advertising.
the author welcomes and appreciates links to and citations of his work. however, if you are viewing the full text of this article at any other website, be aware that its author does not endorse and is not compensated by any advertising or access fees you may be subjected to outside the original web and rss sites.
Interesting post. Can you share some information on how big the data sets you used MERGE for were? In my attempts to use it to date I've found its performance to be less than ideal; between twice as slow and 100x as slow as my pre-MERGE methodology (using a temp table).
BTW, indexed views were introduced in SQL Server 2000 :-)
Also, regarding the idea of creating a "driver table", based on just-inserted keys, did you consider Change Tracking? It might be interesting to work something up using the CHANGE_TRACKING_CONTEXT feature, where the parent process figures out the most recent change version, then generates a unique context and does its work. Then the child process can take the version and the context and figure out what work the parent process did based on the change data.
Can you please elaborate (perhaps in another blog) on your "CTE with a MERGE and a TRY-CATCH block" code? I have done something similar in SQL 2005 before, and am curious how you approached it.
Hi Ward. I believe indexed views were introduced in SQL 2000.
Great post, Ward! My primary server will be upgraded to 2008 in the next couple of months, and MERGE is one of the features I'm most looking forward to (after filtered indexes, of course!).
Also, I've just recently started using indexed views myself and was wondering if you could expound on the improved performance you received in the 1NF table vs the indexed view.
Thanks for sharing. I just starting online.
Its really need. What do you think about www.site2you.com? Is that really possible to make a website pretty quick and is that good quality CMS does they using?