Calculated Dimension ( Permormance Issue)

[krish2459]

Hi,

I'm using the below calculated dimension but for 10 thousand rows the table is very very slow.

Could you please help us to perform the loagic better.

If([Product A Code] = GetFieldSelections(Finished Good Code)

or

[Product B Code] = GetFieldSelections(Finished Good Code)

or

[Product C] Code = GetFieldSelections(Finished Good Code)

or

[Product D Code] = GetFieldSelections(Finished Good Code)

or

[Product E Code] = GetFieldSelections(Finished Good Code)

,[Product E Code])

Thanks.....

Nsc#L2459

[krish2459]

Hi ,

Any suggestions please.

Thanks..

[marcus_sommer · Accepted solution]

Your example indicates that you used crosstable-structures or even more worse a link-table datamodel. Although both methods are technically possible they could have serious disadvantages. If this is the case I suggest to follow the official recommendation to develop the datamodel in the direction of a star-scheme and avoiding/minimizing crosstable-logics.

After all it would mean that you would have a single product-code field and if necessary another field which contained the type respectively the source, like A, B, C ... The usability and also the performance increased usually significantly against the above mentioned approaches.

Beside this you may try to transfer the conditions from the dimension into the expression with something like:

sum({< [Product A Code] = p([Finished Code Code] > + 
            < [Product B Code] = p([Finished Code Code] > +
            < [Product C Code] = p([Finished Code Code] >} Value)

- Marcus