Actions Cube and some attributes

[Former Member]

Hi,

Can anyone tell me how certain attributes like personnel area,sub area, employee group, sub group etc are updated into actions cube 0PA_C01.

Are they updated from update rules of 0HR_PA_0: Employees

or

0HR_PA_1: Personnel Actions

or

Both.

I am interested in this because, one infosource updates monthly and other daily.

The documentation

http://help.sap.com/saphelp_nw04/helpdata/en/b2/f38137ffe27644e10000009b38f842/content.htm is not clear.

Thanks,

Ma

[Former Member]

Hi again,

Yes, the 0CALDAY is populated with the action date on 0HR_PA_1, and therefore the records are stored in the cube on that date and can then be reported with the accuracy of a single day, where as the headcounts are reported with the accuracy of a month.

If you run full loads, you can use a step in the process chain either to delete the cube BEFORE the new loads or to delete old packages AFTER the new loads. Deletion of entire cube is faster, however, if you're process chain fails, then you'll end up with an empty cube in the morning when HR people get in to the office. It's safer therefore to delete the old packages after succesful extraction.

If you have a lot of history data in your SAP HR and the volume is big in a full load, you might want to load all the history data of previous years/months in one full load and then start loading (and deleting) only the data in the current month, for example.

In current project we use abap selection for the infopackage of 0HR_PA_0 to extract sy-datum(6) data and in the process chain we delete previous info package with same selection criteria.

The 0HR_PA_1 is trickier. You cannot load only data for current month, if HR people are doing changes that are dated on previous months. It's a business decision to agree on. We have extracted full load for previous years and with another package we extract data which is restricted with the following abap selection: 20060101 - sy-datum.

Be careful not to use manual restriction for any future months with the 0CALMONTH in 0HR_PA_0, because you will end up getting a lot of unnecessary records. The system will end up sending records for all the future months that you manually include in selection.

Don't try this at home! I once made a manual selection 200501 - 250012 and got over 100 million records...

Regards,

Lauri

[Former Member]

Hi Ma,

Those characteristics are updated from BOTH update rules. It's true that the other datasource brings in monthly values and the other one daily values, but it doesn't cause a problem in the reporting.

The data from 0HR_PA_0, which updates monthly values, can be loaded daily, and it reports the last posted values in every month. Also, the 0CALDAY is derived in the update rules to be the last day of each month.

The other data source (0HR_PA_1) is bringing daily values with 0CALDAY and of course therefore also monthly values in the update rules (0CALMONTH).

Normal reporting would be done with 0CALMONTH, but the actions can also be reported by date. There is no point in reporting Employee counts daily, because it would always be the last date in each month.

Regards,

Lauri

[Former Member]

Thank you Lauri.

Ma

[Former Member]

Lauri,

From one of other threads, it was made clear that CALDAY from 0HR_PA_1 represents action date not the system date.

Moreover both datasources are full loads.

Don't you think, if we load everyday( these data sources), the data will be voluminous in the cube?

or

should we drop the data in the cube completely and load these two data sources everyday(will this have impact on correctness of data in reports?)

Thanks,

Ma

[Former Member]

Hi Madhavi,

It is correct that CALDAY is based on action date.If personnel action has been taken on 15th,and data load is occuring on 16th,then if we use system date CALDAY will show wrong in cube as 16th.

I think both datasources have month and day respectively as selection criteria in the infopackages.

Using the selection criteria u can load data only for that month or that day.

This will avoid deleting complete data target contents,everyday.

cheers,

Vishvesh