Question:
Hi
I am trying to design InfoCube . I need to load the data from 7 various SAP systems data as transaction data . and loading into one InfoCube .
Please give me your valuable suggestions that . I am assuming below options
1. Create seperate InfoSource for each source systems and seperate InfoPackages for each Source systems and create seperate ODS object to each InfoSources and then finally assign all 7 ODS objects to single InfoCube .
2. Single InfoSource and then assign all source systems to the same InfoSource and create seperate InfoPackages for each source systems . and create single ODS object and then assign this ODS object single InfoCube .
please advice me which is the right one .
Thanks
Peter5
Answer:
1. Create seperate InfoSource for each source systems and seperate InfoPackages for each Source systems and create seperate ODS object to each InfoSources and then finally assign all 7 ODS objects to single InfoCube .
2. Single InfoSource and then assign all source systems to the same InfoSource and create seperate InfoPackages for each source systems . and create single ODS object and then assign this ODS object single InfoCube .
my preferred option is the second one;
but it's not mandatory to use an ODS if your scenario don't need it
Ch
Answer:
Hello Peter5,
I just want to confirm CHC's opinion.
Best Regards
Juergen
Answer:
Hi Juergen & CHI
Thanks for your help, I am wondering SAP 3.1i version will support
Delta extraction, If I use LIS extraction or LO cockpit extraction . (SD module) .Is Delta management is based on SAP version or Plig-In version .
Thanks
Peter5
Answer:
I also prefer the second scenario if possible (extracting from CO-PA would be better with scenario 1)
_________________
Juergen Limbach
Answer:
delta management is part of the scope of the PI, but the level of PI that you have also depends on the R3 source system level
if you use LIS extraction (meaning, using the "old-fashioned" way of extracting data with the help of transparent LIS tables), I am pretty sure delta management can be handled as the concept is really basic;
but I am not sure wether LO-COCKPIT concept is already usable in version 3.1i
Ch
Answer:
Hi Juergen & CHI
Thanks for your input again
we need to get below fields from SAP R/3 systems from SD module into
BW system. here are the fields required from two different tables .
still I am at investigation stage that ,what extraction procedure I should use . Do you think LIS extraction is the best method . since we are at
SAP 3.1i .
For your reference I am giving below required fields which I need to extract
1. Company code, VBRK-BURKS
2.. Invoice Date, VBRP-FBUDA
3. Invoice number, VBRP-VBELN
4. Invoice Line Number, VBRP-POSNR
5.Material Number, VBRP-MATNR
6. Invoice Quanity, VBRP-FKIMG
7.Unit of measure, VBRP-VRKME
If I extract from SAP 3.1I version , is it compatable for Delta extraction.
so 'CH' you said in earlier reply that LIS is comptable for delta extraction .
and "Juergen" is also suggesting CO-PA Extraction . can I have best
one please . thanks in advance for your help.
Thanks
Answer:
So you only need thsi fields? Maybe a generic extractor would fir based on a view for the required fields? Can't say anything about extracting LIS because I'm on 4.6 C and use the logistic extraction cockpit.
But tried generating a VIS datasource once, was quit easy to handle.
_________________
Juergen Limbach
Answer:
It depends on your current situation and what skills you have or you can use
1) if you have somebody that is good at configuring LIS in the source system and can customize the system to produce an LIS table taking care about those fields, then you only have to setup the lis environment, do an initial load and then delta uploads ; the "bw" part of the story is REALLY easy to handle, the only problem is to have the custom LIS table setup
2) if you don't have such skills at your disposal, you can create -as Juergen proposed- a generic extractor based on a view ; the problem in this case is to handle the delta management ; I am pretty sure that you cannot have PI 2002.1 in a 3.1i system, so in fact you don't have delta possibilities, it means you will have to do full loads each time (even if you can have a pseudo delta using an ODS, you still have to extract everything)
3) if you have somebody that is not too bad at abap programming, you can use an infoset query fed by an abap program, and then you could handle the delta "manually" in the program ; the rest of the story is the same as for a view
my preference in your situation would be to try 1) first, then 3) and only choose option 2) at the latest stage (but this is because I come from an abap background)
'hope this helps
Ch
Answer:
Hi Ch & Juergen
Thanks to both of yours .
CH : As per your suggestion I feel LIS is most comfortable to work .
you said the "only problem is to have the custom LIS table setup"
1. Is it some thing like required fields selection from LIS InfoStructre.
I hope you might be seen required fields which I want it .
Can I select required fields from LIS Infostructure .
2. As proposed from SAP If we installed (PI 2002.2 (modifying) -R/3 SP A9) in Source Systems SAP3.1i versions . so do you think this would be
fine . am I right . Which Infostructre would you prefer for LIS based on my required fields .
Once again thanks both of you .
Regards
Peter
Answer:
Hi Ch & Juergen
One more questioin I need to extract 0SOURSYTEM field also .
while creating the reports from Cube . I need give in selection option
source system. so I need load 0SOURSYSTEMS . as I said I need
to load data from 7 SAP systems.
1. Is it some thing like I need to extract 0SOURSYTEM from SAP system.
otherwise that field data comes automatically .
2. Apart from SAP systems if I need to load the data from Legacy sytems also ( i.e CSV files) If I am loading data from SAP & Legacy sytems
which field I need to include in InfoSource , is it 0SOURSYSTEM field
is good enough for Legacy and SAP systems or Is it I need to include
0LOGSYS.
Thanks
Peter
Answer:
Well, afaik these fields (if at all neccessary) are filled by BW itself while transfering the data with the logical name of the according source system (*.csv) is a source system also.
Therefore I don't think you have to put attention on these two fields except you want to have queries based on it.
Correct me if I'm wrong...
_________________
Juergen Limbach
Answer:
Hi Ch & Juergen
Thanks to both of yours .
CH : As per your suggestion I feel LIS is most comfortable to work .
you said the "only problem is to have the custom LIS table setup"
1. Is it some thing like required fields selection from LIS InfoStructre.
I hope you might be seen required fields which I want it .
Can I select required fields from LIS Infostructure .
2. As proposed from SAP If we installed (PI 2002.2 (modifying) -R/3 SP A9) in Source Systems SAP3.1i versions . so do you think this would be
fine . am I right . Which Infostructre would you prefer for LIS based on my required fields .
Once again thanks both of you .
Regards
Peter
Answer:
Hmmm, i'm afraid i have to disagree on some points with Juergen and CHC:
- If possible i would go for a generic extractor: invoices can never be deleted from the VBRK/VBRP (in standard SAP), so you can easily create a delta yourself based on AEDAT / ERDAT (changed on/created on)
- I woulg go forn scenario 1: it gives you much more flexibilty in testing/storing/rebuidilding to have your data in an ODS and PSA then just in the PSA.
Answer:
so you can easily create a delta yourself based on AEDAT / ERDAT (changed on/created on)
.
... if you have generic delta option available, that exists only since PI 2002.1 which is not yet available at his premises ...