11

I have a cube which has few big dimensions. It is currently taking 2 hours to Process Full. I have now partitioned the cube and followed below strategy:

  1. Process Update dimensions
  2. Process data Partition (only needed partition)
  3. Process index cube

Because I had few rigid relationship in some dimensions so I wan't able to do Process update. I made all of them as flexible now. But after that processing time of my cube increased to 2 hours 40 min.

Now, my question is why Process update here is taking such a long time? and how can I make it process faster? I am just processing 2 partition at a time out of 34. Here is the breakdown:

  1. Process update all dimensions - 2 hours 20 min
  2. Process data 2 partition - 10 min
  3. Process index - 10 min
Yahfoufi
  • 2,220
  • 1
  • 22
  • 41
Zerotoinfinity
  • 6,290
  • 32
  • 130
  • 206
  • 1
    First of all, please check which dimension is the bottleneck.Please run the processing on dimension using SSIS or Management Studio and save your results. You will be able to see the processing time for each dimension. Than please check if you have no problems with your data and what is your error configuration: do you have a problem with unknown members? – sdrzymala Mar 12 '16 at 15:31

1 Answers1

0

First I would try to delete unnecessary Dimension Attribute, Additionally it is possible to save processing time by setting the AttributeHierarchyOptimizedState to NotOptimized for Attributes that the users don't use often.

Kfir Dahan
  • 1
  • 1