3

How can I aggregate a collection of cubes which all share the same structure onto a central cube with the same structure?

Currently each local server already have a database and a cube, all with the same schema and cube structure.

Each database have a servers table which only contains one globally unique element that identifies that server, and the data on the cube can be discriminated by that ID, so one obvious solution is simply to UNION all interested tables and process a global cube from that.

But since we already have aggregated local data on each server, I'm thinking that can be possible to process a cube with the input from all other local cubes

Is there a way in SSAS to process a cube that is formed by other child cubes?

enter image description here

this question is similar but not equal to this

Community
  • 1
  • 1
Rafael
  • 2,642
  • 2
  • 24
  • 30

1 Answers1

2

You are describing case for SSAS Remote Partitions. Here you can find more detailed description with steps to create Remote Partitions.
Remote Partitioning is seldom used, although it offers a way to scale-out your SSAS installation. Bad news - MS announced this as deprecated feature in SSAS 2016. So the next SSAS probably will not support it, SSAS 2016 still does.

Ferdipux
  • 5,116
  • 1
  • 19
  • 33
  • Thanks. Because of the deprecation the solution would be then to create a centralized data warehouse and feed it with data from all the servers? Or what would be the best way to do it? – Rafael Sep 08 '16 at 18:57
  • @rafael, deprecation means that this feature might not be available in next versions of the SQL. Which means that you have at least 4-5 years of productive usage of SSAS 2016; next version of SQL-SSAS might have some performance improvements to fit all data on one server. So, if from performance or sizing point of view you need Remote Partitions - go for it, understanding maintenance overhead. – Ferdipux Sep 11 '16 at 06:58