Hi

I have cube with total number of rows in fact table about 500K and it got 11
dimensions. Last four month it used take about 25 min. to process. But last 2
days it is taking about 3 hrs. nothing changed same data and cube structure.

What are the things i have to look into to optimize processing,

thanks
kalyan

like a post in the last days...
do you use AS2000or 2005?
have you defrag your indexes in the source database?

does its the loading time (step 1) or the aggregation time (step 2) which
take more time?
1. AS execute a query to load the data; slow performance can come from the
source database
2. AS aggregate the data; slow perf. involve possible memory issue or bad
aggregation design



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

Thanks
I am using AS2000, Slow occured at the time of loading data,
Can u suggestion which best index and also I defrag indexes on fact table.

Thanks
Kalyan

"Jéjé" wrote:

start by updateing the statistics of the source tables
using a command like : sp_updatestats

you can also defrag the indexes using the DBCC commands
search in the SQL help to learn more about these commands.

if the statistics of SQL Server are not correct, the queries executed can
take a very long time because SQL Server will use a bad index plan.
This can explain why your loading time is slow now and not before.

if you delete and add large number of rows in a table, make sure you update
the statistics after this process.


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

simple solution : create a DSO script
or use the cube process utility (found on the MS web site)

"Yolanda Cobos" <ycobos (AT) iniker (DOT) com> wrote