Connecting an application’s data model and queries to a database, including the differences in supported queries, types and schema management between database implementations. Changing an existing schema belongs to schema evolutionSchema evolutionGetting a model change into a database that already holds data, in every tenant, without losing any of it. Generating the initial schema and migrating one that already holds data are often treated as the same step.SAP and community, narrow by design.
Which database the model actually runs on, and what changes when it is not the default one — the choice that is invisible until a query behaves differently. HANA Cloud is the default under a CAP application: HDI containers, calculation views, and the features the core reaches for first. PostgreSQLPostgreSQLA relational database. For CAP it is the production database when the answer is not HANA, reached through an adapter that translates CDS queries into its SQL. through @cap-js/postgres@cap-js/postgresThe PostgreSQL database service — the production option for running CAP off HANA, sharing the base service with the other three.sap official, cap is the alternative when the answer is not HANA, with its own schema evolution and no HDI.
@cap-js/postgres The PostgreSQLPostgreSQLA relational database. For CAP it is the production database when the answer is not HANA, reached through an adapter that translates CDS queries into its SQL. database service — the production option for running CAP off HANA, sharing the base service with the other three.
@cap-js/sqlite The SQLite service, in memory during the inner loop — which is why a CAP project runs before anyone has provisioned a database.
@sap/cds Defines cds.DatabaseService, the technical service every database package subclasses, and auto-wires the installed one, SQLite, SAP HANA or PostgreSQL, for the active profile.
com.sap.cds:cds-feature-hana The CAP Java feature that wires a Maven application to SAP HANA CloudSAP HANA CloudThe managed database under most of this: CAP's default target, the store Datasphere and Analytics Cloud sit on, and a service with plans and components.sap official, btp, the production default among CAP's supported databases.
com.sap.cds:cds-feature-jdbc Connects the CAP Java runtime to JDBC databases through the CDS4j JDBC runtime: H2 and SQLite, which capire names for local development and CI tests rather than enterprise use.
com.sap.cds:cds-feature-postgresql The CAP Java feature that wires a Maven application to PostgreSQL as its datasource beside SAP HANA, with limitations capire lists: schema evolution first, and no multitenancy or extensibility yet.
com.sap.cds:cds-services-bom Creates a CQN-based Persistence Service from the configured datasource for SAP HANA, PostgreSQL, H2 or SQLite, and serves it as the database client.
Not covered: At those four. Anything else — DB2, MySQL, MongoDBMongoDBA document database. Reached from CAP through an adapter over Mongoose rather than through the SQL database services., Snowflake — is an adapter somebody wrote against the same service interface, and the further the store is from relational the more of the interface it has to reinterpret.
On BTP the database is a service you provision rather than a dependency you install, and the choice is made before any model exists. SAP HANA CloudSAP HANA CloudThe managed database under most of this: CAP's default target, the store Datasphere and Analytics Cloud sit on, and a service with plans and components.sap official, btp is the default and what everything analytical above it sits on; PostgreSQLPostgreSQL on SAP BTP, hyperscaler optionA managed PostgreSQL from the underlying hyperscaler — AWS, Azure, Google Cloud or Alicloud — consumed through a BTP service binding.sap official, btp is the managed alternative from the underlying hyperscaler.
Present, not yet described.
Not covered: At provisioning. The platform offers the databases and the plans; which one a model runs against, and what changes when it is not HANA, is decided in the application and read on the CAP door.
The providers here are SAP’s own: managed databases such as SAP HANA CloudSAP HANA CloudThe managed database under most of this: CAP's default target, the store Datasphere and Analytics Cloud sit on, and a service with plans and components.sap official, btp and SAP ASESAP ASE ServiceSAP ASE database systems and tenant databases on BTP with SAP running the platform operations, used from Java applications on BTP or on premise over SQL, JPA and JDBC.sap official, btp, and the HANA drivers for Node.js, hdbhdbSAP's HANA driver for Node.js, written entirely in JavaScript so installing it pulls in no native client, running SQL against SAP HANA and SAP HANA Cloud with streams and prepared statements.sap official, btp and SAP HANA client@sap/hana-clientSAP's own Node.js driver for SAP HANA: connect, run SQL and prepared statements, synchronously, through callbacks or as promises, over native client libraries npm installs per platform.sap official, btp. Core-only reads the catalogue and SAP’s drivers, and is not a quote from the core.
No community artifact is documented here.
On ABAP the database a client connects to is the ABAP system itself: the ODBC driver for ABAP lets an outside program run SQL against CDS objects the system has published.
The ODBC driver for ABAP runs SQL from an external program against CDS objects an ABAP system exposes as an SQL service: queries from release 2021 FPS01, procedure calls from 2025 and inserts and deletes from 2508.
Not covered: At the exposed objects and the client machine. A program reaches only what the system publishes as an SQL service, and the driver is downloaded from SAP and installed on every client, on Unix under unixodbc and no other driver manager.
Core-only rests on the ODBC driver for ABAP, which odbc-driver-for-abap-examplesodbc-driver-for-abap-examplesSAP's example programs for the ODBC driver for ABAP, one subfolder per programming language, with the ABAP objects they use and a unixodbc configuration template for Unix clients.sap official, abap demonstrates, and on SAP HANASAP HANASAP’s relational database platform, which stores application data and executes SQL and SQLScript in the database, available on-premise and as the managed service SAP HANA Cloud.sap official, btp, which ABAP manages and calls through AMDP.
No community artifact is documented here.
The connector itself does not decide the licence boundary. Section 1 makes the destination and purpose of extraction relevant, including the scoped exception for interoperability with an SAP product.
SAP Developer License Agreement 3.2 CAP →Neither the package manifest nor the repository tree declares a licence.
Neither the package manifest nor the repository tree declares a licence.
Neither the package manifest nor the repository tree declares a licence.
Neither the package manifest nor the repository tree declares a licence.
The sqlite:memory preset configures an in-memory SQLite database in all profiles, including production, for edge cases such as an in-memory cache. Before, SQLite was the development and test database and production meant SAP HANA or PostgreSQL.
Advice that SQLite cannot be configured for the production profile at all.
Before cds8, @cap-js/sqlite and @cap-js/postgres were beta previews of a new database-service architecture begun in cds7, and PostgreSQL support in practice meant the community's cds-pg adapter. cds8 concluded that rewrite: @cap-js/sqlite, @cap-js/postgres and @cap-js/hana all reached general availability and became the defaults, sharing one query-translation layer. The 8 line reached npm on 2024-07-12; 8.0.3 was its first published release.
Setup guidance that adds cds-pg for PostgreSQL or treats the @cap-js database services as beta-only. cds-pg was archived in July 2023 in favour of @cap-js/postgres, and since cds8 the @cap-js services are what a new project gets by default.
capability:database-connectors