Load cube 0COOM_C02

Question

Hi experts,
I'm using cube 0COOM_C02, and I'm trying to define the mechanism to load it.
I'm not using any DSO, and the IS I want to load are:
- 0CO_OM_CCA_1
- 0CO_OM_CCA_9
- 0CO_OM_OPA_1
- 0CO_OM_OPA_2
- 0CO_OM_WBS_1
- 0CO_OM_WBS_2
I wonu2019t be loading  0CO_OM_NAE_1,  0CO_OM_NAE_2,  0CO_OM_NTW_1, 0CO_OM_NTW_2, 0CO_OM_NWA_1 and 0CO_OM_NWA_2, as Network* information is not being used in R3.
ODS is not being used
From  sap help I got the following regarding the loading of these IS:
1) Create an InfoPackage for InfoSources 0CO_OM_CCA_9, 0CO_OM_OPA_6, 0CO_OM_WBS_6, 0CO_OM_NTW_2,  0CO_OM_NWA_2 and 0CO_OM_NAE_2 to load the actual data.
Case A (with delta process): If you use the delta process, you can load your actual data at any time. You do not need to enter the period.
Case B (without delta process): If you do not use the delta process, it is best to load your actual data at the end of the period (after completion of the period-end closing process) by restricting the selection to the expired period.
SAP recommends the delta process.
2)To load the remaining data (such as plan, target or period-end closing) for InfoSources 0CO_OM_CCA_1, 0CO_OM_OPA_1, 0CO_OM_WBS_1, 0CO_OM_NTW_1, 0CO_OM_NWA_1 and 0CO_OM_NAE_1, create a separate InfoPackage for each type of data. This is necessary because of the different data transfer time points. Restrict selection to the relevant value type (such as plan) and the period or periods (such as 1 u2013 3 for the first quarter or 1 u2013 12 for the full year).
From your experience in implementing this cubes, I would like to confirm the following:
Q1: I assume that  1) will be used just to load actual data. If I use delta process (which is what I want), I can load data for actual everyday for exemple. Will the remaining cases (2) just be used to target, plan and commitment. Am I correct?
So IP mentioned in 1) only actual data will be loaded, correct? No restriction has to be done in version at IP level?
Q2: Regarding other than actual data (I only have planned), here is my scenario
Cost Center u2013 plan defined once a year (each end year)
WBS u2013 plan defined every time a WBS is created. I will execute reports every month
Order u2013 plan defined every time a order is created. I will execute reports every month
Changing the plan data of a Cost center, WBS or order, might happen
Taking this scenario into account, I came up with 2 possibilities to create plan data loading PC (to be run automatically every day 1 of each month)
IP created
Cost Center u2013 IP1 for 0CO_OM_CCA_1, with version and type restricted (to have plan data)
WBS u2013 IP2 for 0CO_OM_WBS_1, with version and type restricted (to have plan data)
Orders u2013 IP3 for 0CO_OM_OPA_1, with version and type restricted (to have plan data)
OptA
Del Index
Execute IP1 (load data to the cube)
Delete Overlapping request
Execute IP2 (load data to the cube)
Delete Overlapping request
Execute IP3 (load data to the cube)
Delete Overlapping request
Generate index
Q2.1: Iu2019ve seen this option (Delete Overlapping request) is another thread, but I donu2019t know much about it. Can somebody explain how it works and how will affect the loading performance?
OptB
Del Index
Execute IP1 (load data to the cube), routine to restrict by fiscal period (month year)
Execute IP2 (load data to the cube), restrict by fiscal period (month year)
Execute IP3 (load data to the cube), restrict by fiscal period (month year)
Generate index
Q2.2: I think this might work better in terms of performance, but I donu2019t know if and how is possible to create a routine to make it load just last months records. Also, is older data changes, this wonu2019t be an option. Can you advise on this solution?
Q3: I will also be creating a PC to load actual data (delta). Do I have to include both this chains in a metachain, or can I have 2 separate PC?
Many thanks
Regards
Joana

Answer 1

Hi,
Q2: IS 0CO_OM_WBS_1. I just want to extract plan from here. Ideally, i would do a Init + IP for this. But it's not possible. Just full IP: I want to be able to load this every month, without duplicating cube data.
It would be possible to load, by filtering fiscal period (previous month) + plan data. But i'm afraind they might chage older data, and by loading just previous month I would loose those records.
So that's why i thought of always loading every plan data, deleting previous request from cube.
Its depending on the volume of the data, If the data volume less you can delete the last request  and then full load(since it is monthly not a big issue)
*IF the data volume is huge then you can go for selection in the info package(full with previous month) 
Generally once the plan data is published then it would not have changes*
If the data volume is manageable then you can go with full and delete the previous request
By doing that, on PC, will it have any impact on actual data in the cube? NO
Q3: Can I havw 2 PC: one for actual (based on 0CO_OM_WBS_6) and one for plan (based on 0CO_OM_WBS_1)? Or do I need Metachain? I don't think so, because i'm running them in different times (one everyday, the other every month.
2 separate chains require since one is daily and another one is monthly
For this cube: 0COOM_C02, are you aware of another solution to improve loading performance
At this point of time this model will give a ideal performance once data volume is increased the you need performance method
Regards,
Satya