Hi

I have a quite big Data Warehouse with 1 table with 20.000.000 rows in 20
columns and about 5 tables with 40 rows in 2 columns. The cube contains 20
dimensions.

I have a problem with processing the cube. During the processing the cube
after some period of time the server restarts it selves. So, I removed 16
dimensions and started processing the cube again. There was no problem with
processing the cube. After adding again dimensions one by one I managed to
process the cube only with 4 additional dimensions. I still have to add rest
12 dimensions.

I think that server is restarting because processors (2 x PIV 3.0 ghz) are
overheating.

Can you tell me how to tune my cube? I know that 20 dimensions it is a lot -
maybe even too much, but right now they have to be included into cube.



P.S.

1. Reducing number of rows in DW also did not help.

2. These 16 dimensions were created because the client has a 16-characters
product code and he want's to be able to perform some analysis using
combination of these characters.



Regards

first point 20 millions is not so big.
after... make sure that the cube is optimized (search in the menu when you
edit a cube optimize)
This remove the (or at least improve) the usage of joins when AS read the
data source.
(less joins result in faster reading step)

also, have you identified if its an hardware issue?
generally a "real" server has enough fan.
Try to monitor the temp. of your CPU (most of the motherboard provides a
utility to do this job or you can find a generic tool on the web like
http://www.hmonitor.com/ )
try to change bios settings to reduce the CPU speed automatically when the
temperature is too high instead-of stopping the server.


"herman" <rhermaszewskiWYWALTO (AT) o2 (DOT) pl> wrote

Thank you
I will try.

Regards


U¿ytkownik "Jéjé" <willgart (AT) BBBhotmailAAA (DOT) com> napisa³ w wiadomo¶ci
news:%23GwG6baPGHA.2624 (AT) TK2MSFTNGP12 (DOT) phx.gbl...