Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
SqlZ
Yak Posting Veteran
69 Posts |
Posted - 2001-11-09 : 09:03:17
|
| Just a little backgroud on what I have and what I am trying to do. I am designing a denormalized DB to resemble a data warehouse cube with one Fact and 3 Dimensions to eventually be used as a cube in Analysis Services. My fact table is huge, ~34 million recs and I need to load the supporting Dimension IDs into the fact. I have turned to cursors and they are crawling. Updating 3000 recs/min. I have tested the performance on a RAID 5, P3, 700 Mhz, 1GB RAM Real Server, and my Sandox Server, which is a souped up Desktop PC that is P3, 640MB RAM, 450 Mhz, and it peforms twice as many rows per minute as my real server. I have dedicated about 3/4 of each systems' physical memory to SQL Server in both instances. Any ideas on how I can bump up the performance and optimize these servers? Thanks in advance.MCDBA, Senior Engineer, DW Architect. |
|
|
|
|
|