Summary
The IBM Db2 Haystack integration currently only supports synchronous
pipeline execution. This issue tracks adding full async support using
the AsyncConnection and AsyncCursor APIs introduced in ibm_db 3.3.0,
so the integration can participate in await pipeline.run_async(...) calls.
Motivation
Haystack's Pipeline now exposes run_async() / run_async_generator().
Components that implement async def run_async(...) can participate in
async pipeline execution. The IBM Db2 integration is currently sync-only,
which blocks users who want to run Db2-backed pipelines asynchronously
alongside other async components.
ibm_db 3.3.0 (released) introduces a native asyncio API
(AsyncConnection, AsyncCursor) that wraps synchronous IBM CLI calls
in asyncio.to_thread(), making it possible to add async support without
any additional runtime dependencies beyond Python's standard asyncio.
Summary
The IBM Db2 Haystack integration currently only supports synchronous
pipeline execution. This issue tracks adding full async support using
the
AsyncConnectionandAsyncCursorAPIs introduced inibm_db3.3.0,so the integration can participate in
await pipeline.run_async(...)calls.Motivation
Haystack's
Pipelinenow exposesrun_async()/run_async_generator().Components that implement
async def run_async(...)can participate inasync pipeline execution. The IBM Db2 integration is currently sync-only,
which blocks users who want to run Db2-backed pipelines asynchronously
alongside other async components.
ibm_db3.3.0 (released) introduces a nativeasyncioAPI(
AsyncConnection,AsyncCursor) that wraps synchronous IBM CLI callsin
asyncio.to_thread(), making it possible to add async support withoutany additional runtime dependencies beyond Python's standard
asyncio.