[Mondrian] Fact table having invalid join key with the dimension table

Luc Boudreau lucboudreau at gmail.com
Tue Mar 4 12:08:51 EST 2014


Mondrian assumes that all keys are valid and present. If I may ask, why
have invalid keys in your fact table? If there are no reasons for them, I'd
recommend to simply fix your database.

Luc
On Mar 4, 2014 11:52 AM, "Mayur Mohite" <mayur.mohite at vizury.com> wrote:

> Hi,
>
> We had few issues dealing with rows containing invalid join key in the
> fact table.
> Our fact table F has rows with invalid join key with dimension "Geo".
> When we aggregate the fact table by date dimension on a particular date
> say "2014-03-03" and if we choose another dimension say "Geo" and query for
> all members of "Geo" (Geo.members) we get aggregated numbers whose total
> sum ("discarding" Geo members and summing up the returned result) is not
> equal to the aggregated numbers when we query mondrian for "2014-03-03"
> alone (due to fact table having invalid join key entries with Geo).
>
> After this query if we aggregate the numbers for the date "2014-03-03"
> alone, mondrian fetches the result by summing up numbers from the cache
> warmed up by the previous query. Due to which the resulting aggregated
> numbers are lower than actually present in the DB.
>
> Is it possible to instruct mondrian to not populate the cache for the
> queries involving Geo dimension (Fact table has invalid join key of that
> dimension) via Schema XML or via mondrian.properties file?
>
> -Mayur
>
> _______________________________________________
> Mondrian mailing list
> Mondrian at pentaho.org
> http://lists.pentaho.org/mailman/listinfo/mondrian
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20140304/29e2c7e2/attachment.html 


More information about the Mondrian mailing list