Is there is a way to create cube against database which tables get new records every few seconds?
I get these errors
Thanks in advance
Warning -2128674815 : Errors in the OLAP storage engine: The attribute key cannot be found: Table: dbo_Log, Column: InsertDate, Value: 2/23/2007 9:22:32 AM.
Thanks in advance
Hello. I have not build anything like this but my guess is that you will have to use ROLAP for dimensions and Measure Groups.
The error you see probably depends on that you have a MOLAP cube where the dimensions have not been processed before the partitions/measure groups.
Keep your datamart in the relational engine is my recommendation. Five seconds is a very short batch window. I suspect that you have a cube directly on top of an OLTP-system.
Maybe Notification Services is a better choise than SSAS2005?
Regards
Thomas Ivarsson
|||You'll need to update the time dimension and your data incrementally and "lock" the database updates while the update is being performed (to maintain consistency). This'll take some careful consideration of your partitioning scheme, to make sure you maintain a reponsive system.
How often do you need to update the cube? Could you consider doing the following steps:
a) materials the dimension information into a "Time" table
b) use a view "dbo_Log_view", that joins to the "Time" table to only return records that exist in the "Time" table. (Alternatively, you could do the same on the partition SQL statement, but I recommend abstracting it to the view.)
c) Process the cube / partition incrementally?
Note: You could set this up in a SQL Agent job, to run as a two step process automatically.
|||
Thank you guys.
I will have to carefully review all the options. Thank you for all suggestions.
Best Regards,
|||I think you need Integration Services for this. I haven't figured it out myself yet, but I need to make this too, and I came across functions in SSIS wich seemed to be usefull to accomplish this.Regards,
Eyso
No comments:
Post a Comment