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 |
|
influent
Constraint Violating Yak Guru
367 Posts |
 Posted - 2012-08-14 : 20:37:26
|
| My company would like to show our customers analytical data in very close to real time, things like how many orders they've ever placed, the statuses of their latest orders, etc. If we were to use the production database for this it would require lots of indexes and massively increase CPU usage, etc., so that's not an option. We have a replicated copy with lots of indexes that we use for reporting with SSRS but (a) I still think performance is not good enough and (b) our schema is constantly changing in ways which seem to make it impossible for us to do deployments without dropping the publication first and having to wait an hour for the recreated publication and subscription to get sync'd. I'm wondering if a cube in SSAS is the answer, using incremental data loads in SSIS based on the rowversion column in each table. I'm just not sure it can check several tables for updates, push the updates to the data warehouse, and process the cube all within 5 seconds.Thanks in advance. Any input would be appreciated. |
|
|
|
|
|
near-real-time data loading into SSAS from SSIS?