The creating aggregations of processing is very slow on one of my cube. The
performance monitor shows the "Temp file bytes written/sec" is 10MB/s - after
I increased the "Process Buffer Size" to 200M.

Sounds the temp file written caused the problem.
Or anyway to reduce the memory consumption of dimension data? I want also
create more aggregations - currently some partition cannot have more then 40
aggregations, otherwise memory error.

Creating the 4th aggregations starts to use temp file when I set Process
Buffer Size to 200MB. And creating the 7th aggregations start to use temp
file when I set the buffer to 400MB. So it seems 3 aggregations will use
200MB.

And it cause memory error (Server memory low) when creating 30 to 40
aggregations.

The cube has 14 dimensions. Any skill to make it use less memory?

"nick" wrote:

Since the cubes have 14 dimensions which is bit bigber cube
try
setting Aggregation usage property of the each dimension to "TOP Level
Only"
and use "Performance gain reaches" between 15 to 20 % during design the
storage for the cube.

Regards,
Prasanna

One possible source for (server memory low) is if you are running out of
system page file. The default by Windows is not alway appropriate for the
way AS uses memory. If you look in the AS Operations Guide, we recommend
increasing the default size up to the maximum and adding a second page file
(sized with the amount of physical memory on the machine) on another disk.
http://www.microsoft.com/technet/pro.../anservog.mspx

--
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.


"nick" <nick (AT) discussions (DOT) microsoft.com> wrote

The OLAP server already installed 4GB RAM. So the page file should help since
the physical memory already cover all 32 bit address space? However, for
policy reason, we cannot enable the /3GB switch yet.

"Dave Wickert [MSFT]" wrote: