I see two ways - not sure about pro's and con's

1) Pulling out data from one fact table and creating new fact tables using
SQL
2) Using source table filter in AS without doing anything in SQL

--Michael V.

If #2 is an option that I'd got for that every time! The tmie and effort it
saves you in ETL more than makes up for the minor load increases that you MAY
put onto your fact table.
If you don't have users accessing your DW directly then its a no-brainer
(IMO).

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


"Michael Vardinghus" wrote:

Thanks Jamie

Tried option 2 and that was a no go...don't get the idea of this filter but
you cannot filter dimension values that exists in the fact table...it will
give an error when processing cube.

About 1 - what to choose:

DTS or replication...is replication faster ... ? I don't need to wash
because I'll just take a copy of data in the fact table but my fact table is
emptied every night and filled again with all records - does that mean that
I wont get the replication benefits ?

"Jamie Thomson" <jamie.thomson (AT) donotspam (DOT) conchango.com> skrev i en
meddelelse news:0B858428-8EA2-4D4E-A2F7-C481BDA46C30 (AT) microsoft (DOT) com...

Why won't option 2) work? I thought you were putting the view(s) onto the
fact table, not the dimension tables.

DTS or Replication. I personally would go for DTS but thats cos I know alot
about DTS and nothing about replication.

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


"Michael Vardinghus" wrote:

Of course it will...my mistake....

I put a view on top of the fact table ....

Thanks Jamie


"Jamie Thomson" <jamie.thomson (AT) donotspam (DOT) conchango.com> skrev i en
meddelelse news:86865323-4417-49A6-9DC8-AABC20596B67 (AT) microsoft (DOT) com...