Intercube feeder

[sbalaz]

Hi everyone,

is there a way how to workaround with intercube feeders in cubes which seems to be too large? I have got cube A (17 dims) and cube B (17 dims too).

There is following rule in B loading data from A:

Code: Select all

['Local currency','No Week','Actual','ActualSAPdata']=N:
    DB('SAP Actuals II','Total Salestypes',!Location SOC,'Total EMEA + Other + IG_Elim',!Accountmgr,!Currency,!Country,'SAP 3000',!Week,!Period,!Fiscal year,!Product,!SalesOrderItem,'Local currency',!Customer POR,!EndCustomer POR,!SalesOrder,!Sales measure)\1000+
    DB('SAP Actuals II','Total Salestypes',!Location SOC,'Total EMEA + Other + IG_Elim',!Accountmgr,!Currency,!Country,'SAP 2093',!Week,!Period,!Fiscal year,!Product,!SalesOrderItem,'Local currency',!Customer POR,!EndCustomer POR,!SalesOrder,!Sales measure)\1000;

I tried following feeder in A cube (A=>B):

Code: Select all

['Total EMEA + Other + IG_Elim','Local currency',{'SAP 3000','SAP 2093'},'Total Salestypes',{'booking','sales'}]=>
  DB('SOC Details',!SalesGroup,!Accountmgr,!Currency,!Country,'Actual','No Week',!Week,'ActualSAPdata',!Period,!Fiscal year,!Product,!SalesOrderItem,'Local currency',!Customer,!EndCustomer,!SalesOrder,!Sales measure);

The problem is that this consumes too much memory and brings the server (4GB RAM) down when saving the rule/feeders for A cube.
I have managed to save the rule/feeders for cube A before it was populated. However when loading data to cube A via TI process from the text file, it brought the server down.

I have got similar/the same rule and feeders between two cubes (let us say A_old and B_old). Additional information is required by users (Sales Order, Sales Order Item, Account Manager and Product), therefore I needed to create new cubes. The only differences are:
1. Two dimensions are not included in A_old cube: SalesOrder (cca 15000 leaf elements) and SalesOrderItem (cca 300 leaf elements).
2. Four dimensions are not included in B_old cube: SalesOrder (cca 15000 leaf elements), SalesOrderItem (cca 300 leaf elements), Product (cca 300 leaf elements), Accountmgr (cca 300 leaf elements).

I am not sure if this is just too much data for TM1 server, or I am overfeeding the cube. Is there a way how to workaround with this? Tighten feeders even more, use a lookup cube for sales orders or something else...

Thanks for help.

Stanislav

{Admin note: Moved from the Enhancements forum. This is not an enhancement request. Please read the How To Make An Enhancement Post thread before posting in that forum.}

[Marcus Scherer]

Hi Stanislav,

since you are transfering Actuals, you could as well transfer the values with TI and store them physically in cube B. Assuming you have no end user interactivity for Actuals. Write a process that runs after you have loaded your Actuals in cube A.
Another point is the summation SAP3000 + SAP 2093 in your rule. You could create a summation element in your hierarchy nad refer to that one in your rule. Same for booking + sales.

HTH for the first part - I don't understand your question with your x_old cubes.

Marcus

[sbalaz]

Hi Stanislav,

since you are transfering Actuals, you could as well transfer the values with TI and store them physically in cube B. Assuming you have no end user interactivity for Actuals. Write a process that runs after you have loaded your Actuals in cube A.
Another point is the summation SAP3000 + SAP 2093 in your rule. You could create a summation element in your hierarchy nad refer to that one in your rule. Same for booking + sales.

HTH for the first part - I don't understand your question with your x_old cubes.

Marcus

TI process for transfering actuals to B cube looks like a very good idea. I will try this way.

Regarding x_old cubes - this was just an examle which is in place and works well. I just wanted to describe differences between working (old) cubes vs. A and B cube.

Thanks a lot.
Stanislav