Loading a warehouse
Getting data into an analytical store and keeping it current — connections to source systems, replication and the other integration tasks, import jobs and their lifecycle, and the monitoring that tells you a load failed before a report does. Distinct from replicationData federationReading data that lives in another system as if it were local — by federating the query, by copying the rows, or by caching what was read. Three different answers to one question, and the choice is about how stale you can afford to be.SAP and community, narrow by design, which is an application reading data that lives elsewhere as if it were local rather than a warehouse being filled.
Overview
Getting data into an analytical store and keeping it current — connections to source systems, replication and the other integration tasks, import jobs and their lifecycle, and the monitoring that tells you a load failed before a report does. Distinct from replicationData federationReading data that lives in another system as if it were local — by federating the query, by copying the rows, or by caching what was read. Three different answers to one question, and the choice is about how stale you can afford to be.SAP and community, narrow by design, which is a CAP service reading data that lives elsewhere as if it were local rather than a warehouse being filled.
The product owns the pipeline and the roles around it: in Datasphere, a space administrator or integrator creates connections to source systems and databases and schedules and monitors data replication and other data integration tasks, alongside importing business content and controlling access to the space.
Not covered: At the tenant's own scheduler and monitor. What ran, what failed and what has to be re-run is visible in the product, and getting that into whatever else watches your landscape is nobody's job by default.
SAP's answer here is a product rather than a service you bind, and the products are now entries — Datasphere, Analytics Cloud, Business Data Cloud and HANA Cloud as BTP services, BW/4HANA as an external because it is installed rather than subscribed. Both sides of this assessment are artifacts.
Tools that reach the loading surface from outside the product: BW's Push API and request monitor, SAC's import-job lifecycle, Datasphere's ETL extraction, and SAP's dbt-sap-hana-clouddbt-sap-hana-cloudA dbt adapter, installed as a Python package beside dbt-core, that lets a dbt project target SAP HANA Cloud as its warehouse through a profile in profiles.yml.sap official, btp adapter, which loads dbt models into SAP HANA Cloud.
On ABAP this is loading SAP BW, where each transformation from a source into a DataStore object may carry ABAP start and end routines, and those routines are where custom cleansing and checks run.
SAP BW transformations carry ABAP routines for rules the mapping cannot express: a start routine runs on each data package before the rules, an end routine post-processes each package after them, and an expert routine covers cases the other functions cannot.
Not covered: At the hook. Each routine is generated into its own transformation's program, so a check every flat-file load needs, such as removing characters an InfoObject rejects, still has to be written into or called from every transformation that runs it.
BW here is SAP BW/4HANASAP BW/4HANASAP's data warehouse product for on-premise and private-cloud landscapes, optimised for the HANA database and installed rather than subscribed — the 2023 release is current, with its own master and upgrade guides. Outside the atlas because there is no adoption decision to model: you have a BW system or you do not, and what is entered here is the tooling that reaches one. or a BW release on NetWeaver; the routine behaviour is read from the BW 7.5 documentation.
Reusable ABAP classes for those routines: one strips characters an InfoObject would reject at activation from an end routine, another attaches a class to a transformation's start routine from the transformation's ID so each one is not edited by hand. ERPLERPLA DuckDB extension that reads SAP ERP tables, CDS views and BW queries from SQL over RFC, BICS and ODP, and can mount an SAP system as a DuckDB catalog to join against other data.independent, abap pulls SAP ERP and BW data into DuckDB through ODP extraction, in full and delta.
Artifacts
Content
Seen working
Details
Coverage
For your agent
- Ref
capability:warehouse-loading- JSON
- /capabilities/warehouse-loading.json
- MCP
- Connecting an agent