Dear all,
we have a situation where we consider persisting some data on ECC (essentially BOM data) and consume that in IBP via Cloud Integration for data services (CI-DS)
AFAIK we have two general options
- Use an extractor to encapsule the persited data (or do not persist at all but compute within extractor)
- directly access the table with the persisted data
Let's assume the persisted table exist and it holds several thousand rows but I'm only interested in those two rows
My question revolve around performance / amount of data fetched / filtered / queried.
- Am I right in the assumption, that a extractor would grant me ways of selecting data (e.g. only specific products) by means of input parameters?
- When using direct table access, will such a selection (A + B) be done end to end or will the whole table be fetched (hence pushed to cloud integration as a whole) and filtered only there to A+B?
- Will an abap query help, in case 2. will normally result in a whole table read?
- What about caching in the general options 1. Extactor / 2. direct table access
Generally spoken I'd like to get a feeling what would be considered best practice when it comes to performance and not pushing unnecessary data to CI-DS when dealing with large (row) tables / data sources (in ECC)
Many thanks and cheers
Jens
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.