Load from Data Stream:Input Type

[Former Member]

Hi,

What does Input Type signifies in under Input Type > Load From Data Stream Method?

If Input type is selected as Cumulative and in R/3 data is available for period 01 and period 02 what will happen if I start consolidation in BCS from period 02 and load data in period 02 only?

Will it pick up cummulative data of period 01 & 02 from R/3 and stored in period 02 of BCS? Please advise.

Load from Data Stream Method has update mode as ´Delete All'.

Best Regards,

UR

[Former Member · Accepted solution]

Hi,

IMHO, the system will pickup data from period 02, calculate the difference between this data and data already loaded into BCS in period 01.

But I would restrict the source in the LFDS by period = 01 & 02 and for the period in mapping use 01 or 02. This would bring your two periods from the staging cube to one period in BCS. Otherwise, if data in the staging cube are kept periodically, then you'll get just one month data.

Edited by: Eugene Khusainov on May 7, 2009 5:51 PM

[Former Member · Accepted solution]

Hi Eugene,

In BCS data has not been loaded in period 01 and it is not intended to be.

Period 02 is the first period of consolidation in BCS. However in source system data is from period 01 onwards.

Does your reply means, that even if Input Type is selected as ' Cummulative', LFDS will select only period 02 data from source system?

Please advise.

Best Regards,

UR

[dan_sullivan · Accepted solution]

In most cases with this scenario that data loaded for period 002 will be cumulative and thus include both periods 001 and 002. However there are some dependencies on the configuration of the source infocube and the load-from-datastream mappings.

[Former Member · Accepted solution]

First of all, you need to find out how data are stored in the staging cube.

If the data is cumulative, just load period 002 into period 002 in BCS. That's it.

If it is periodical, use the restrictions for source and target in LFDS, as I described earlier.

[dan_sullivan]

Yes, if you have input type cumulative and execute data collection for the first time in period 002, then period 002 will include data for both periods 001 and 002 of the source system.

Delete all ensures that when the method is executed multiple times in the same period, the data for that period is deleted before the new data is loaded.

One caution with Delete All is to be sure that the source data includes all cons units if this is to be executed for a top node, or that the execution only occurs for those cons units with data in the source infocube. Otherwise, the data for all units is deleted and only the source data is loaded.