Going back, for now

The SQL 2005 cluster is running really well. I migrated the DTS packages from the SQL 2000 server into the new SQL Business Intelligence Development program. This is the program where I can edit the packages and then deploy them to the SQL Server. There isn’t a DTS environment anymore. It’s now SSIS (SQL Server Integration Services).

Some of them work fine but others require a major rewrite. The SSIS packages have their own looping mechanism so the ones I wrote into a couple of the SQL 2000 DTS packages don’t work at all. I can tell a speed difference on the SSIS packages that actually do work (they’re much faster). I had one DTS package that took almost 5.5 minutes to complete because it had a lot of loops. In the new SSIS package, it completed in 2.5 minutes. Not bad at all.

The problem with some of the existing packages, and the ones I’m going to have to finish, is the reason I’m going to end up releasing the data warehouse in SQL 2000. I have to get it finished within the next month to month and a half. That’s not enough time for me to learn the new SSIS package development (it’s all Visual Studio .NET) environment and finish the work. I can finish the warehouse in SQL 2000 and later move it to 2005.

There won’t be any difference in the tables and everyone uses a BusinessObjects reporting server to compile reports. All I will have to do is change the SQL server to which BusinessObjects connects. Any reports will work the same. It will also have the same SQL user, permissions and default database. No one should notice the change when it is made.

I have to setup the new BusinessObjects server next Monday. It’s also going to be clustered and the interface will go on our existing SharePoint Portal Server 2003. Even though not all the data is in the warehouse, they want to start writing reports based on what is in. I’m short four (4) standardized test groups out of nine (9). One of the ones I’m short I am 90% through writing the DTS package. I only lack one piece which I should get tomorrow or Monday.

Sometimes it’s a good idea to install an upgraded server on a seperate machine instead of inline (same server). If I had done an upgrade of my SQL Server (2000 to 2005), I’d be in a bind right now.

18.May.06 Microsoft SQL