Does anyone have an ETA table of how long a cube will process if you
have "x" number of dimension, "x" number of measures, "x" number of
hiearachies, your processor, your memory, the number of bytes of your
datawarehouse, etc.

A formula of how to calculate this would also be appreciated.

The biggest influencing factor will be the number of rows in your fact table.
There's no formula that will take that and all of the factors that you have
mentioned into account.

Regards
Jamie Thomson
http://www.conchango.com

"Rico Cali" wrote:

The general rule of thumb that I use is that on server-class hardware you
should be able to do at least 1 million rows per minute with no aggregates.
Like all rules of thumb your mileage may vary. I've seen faster (with higher
quality I/O subsystems; faster gigibit networks; etc.). I've see slower with
systems with lots of aggregates. But, in general, that is what I expect to
see.

If you have considerable less performance than that, check to see if you
have ran the Optimize Schema wizard in the cube editor and have simplified
your SQL statement as much as possible; or if your OLEDB provider (or ODBC
driver) is providing as quick a throughput as possible (this sometimes
happens with some of the older legacy data sources with older ODBC drivers).
--
Dave Wickert [MSFT]
dwickert (AT) online (DOT) microsoft.com
Program Manager
BI SystemsTeam
SQL BI Product Unit (Analysis Services)
--
This posting is provided "AS IS" with no warranties, and confers no rights.

"Rico Cali" <ricocali (AT) hotmail (DOT) com> wrote