Dataflow performance

Hi,

We are running a dataflow which is lasting too much, about 50min. Something seems to be wrong but we can not figure it out.

The goal of the Dataflow is to adda new entity to the original cube. The Dataflow multiplies the original cube with a cube (a kind of diagonal matrix) which contains a one to one relation from the RL entity to the new entity we want to add, the RL_Receiver. The calculation domain is extended only to this new entity.

It is something that we have done many time but for some reason this case has lasting a lot.

What could be the cause? The types of the cubes, the sparcity, the calculation domain, the preselections before the Dataflow…

Find more posts tagged with

Accepted answers

Samir Jones

@Jordi Vilamajo, if the Temp Cube contains the original 7 entities and the added entity (total 8 entities), I believe you could use the expression JOIN(a*b). If you try the JOIN, it is important that no entities are extended.

All comments

Samir Jones

Jordi Vilamajo

Hi Samir, that worked perfect. We did not know we could use that function. Thanks

Robert Barkaway

Just be aware that the JOIN() function is deprecated from Board 12.6 onwards, as Board now automatically applies such functions without you needing to explicitly use them, it won't cause any harm but it will be unnecessary and will be ignored by the engine.