Skip to content

Latest commit

 

History

History
72 lines (58 loc) · 5.14 KB

File metadata and controls

72 lines (58 loc) · 5.14 KB

Additional database qualification

Reviewed 22 September 2026 against FluentMigrator's current runner projects and provider configuration. This is an implementation gate, not a claim of released provider support.

Additional engine Real-engine GitHub Actions route Current disposition
SAP HANA Official HANA Express Linux container and SAP's .NET driver, disposable schema Implemented with a mandatory actual-engine matrix job; require green PR checks before merge
Amazon Redshift AWS test warehouse/serverless endpoint with CI credentials, network access and resource cleanup No configured test infrastructure; defer provider
Snowflake Snowflake test account, warehouse, credentials and disposable database/schema No configured test infrastructure; defer provider
Db2 for IBM i IBM i endpoint on Power infrastructure and compatible .NET/ODBC driver No configured test infrastructure; defer provider

The existing Db2 job runs Db2 LUW, not Db2 for IBM i. PostgreSQL compatibility does not prove Redshift behavior. Snowpark's local test framework does not test Snowflake DDL and catalog behavior through the production .NET driver.

Older provider lists also mention SQL Server Compact and SAP SQL Anywhere. FluentMigrator's current FAQ marks these as dropped; they are not current additions to pursue. Different SQL Server/PostgreSQL dialect versions and Oracle drivers are not additional database engines.

HANA admission criteria

The HANA matrix startup pins HANA Express 2.00.088.00.20251110.1 and Sap.Data.Hana.Net.v8.0 2.30.27. It uses a disposable public CI credential, bounded startup, native .NET connection, identity/constraint DDL, persisted data, rollback and catalog checks. Any startup or behavioral failure fails the job. Logs and runner resource evidence are retained; cleanup runs independently of test success. The prerequisite probe passed in run 35766200488. The standalone workflow is replaced by the mandatory Hana job in the complete database matrix; the probe project remains reproducible evidence.

The provider matrix at source eabec55 is recorded in run 35770116342. Provider admission requires this matrix to pass, covering imperative/fluent schema creation, constraint metadata, data, migration history, restart/rollback, preview parity and explicit unsupported operations. Do not mark a provider supported on the basis of SQL string tests or a skipped secret-gated job.

Primary sources

HANA provider scope

The source provider is selected by ProviderTypes.Hana and accepts a caller-owned HanaConnection or the SAP factory. The core remains free of SAP driver references. The tests use Sap.Data.Hana.Net.v8.0 2.30.27 and HANA Express 2.00.088.00.20251110.1.

Supported operations include row/column table creation, explicit keys/checks/FKs, column add/change/remove/rename, table rename/remove, ordinary and unique indexes, parameterized data operations, schema/constraint/index metadata, versioned history, and structured create-table/add-column preview. Names are case-preserving and accept unquoted table or schema.table input; embedded identifier dots require explicit SQL.

Whole-session transactions and native locking are not advertised. The provider retains HANA's default DDL autocommit behavior; data rollback does not prove DDL rollback. Use explicit SQL for tenant administration, computed columns, specialized indexes, collation configuration, and provider-specific data types without a mapped CLR type. The provider rejects unsupported included/filtered/clustered indexes and semantic collation requests instead of ignoring them. The packaged CLI does not bundle the SAP driver; use the library runner in a host that references the SAP package. Raw defaults retain HANA's engine restrictions: CURRENT_TIMESTAMP is valid, whereas arbitrary LOWER(...) defaults are not.

HANA constraints, referential constraints, and DDL autocommit are documented by SAP.