potiuk opened a new pull request, #68814: URL: https://github.com/apache/airflow/pull/68814
The provider DB tests run as parallel test-type groups (`breeze testing providers-tests --run-in-parallel`) but **without xdist** — each group runs serially. The "all other providers" group bundled ~all providers into one `Providers[-amazon,celery,google,standard]` type and became the **serial bottleneck**: it ran ~15 min while the other parallel slots went idle after a few minutes (one slot doing all the work on a 4-core runner). This splits that group into `NUMBER_OF_PARALLEL_PROVIDER_SLICES` **balanced chunks** so every parallel slot stays busy. The chunks are enumerated from the same source as the negation (`_get_individual_providers_list`), so their union covers **exactly** the same providers — a test asserts no provider is dropped or duplicated, and a coverage-invariant test guards it. It also drops `standard` from the isolated "long" providers: its `PythonVirtualenvOperator` tests used to dominate the suite, but they are now DB-free and parallelize under xdist (#68533), so `standard` is no longer a slow group and belongs in the rebalanced chunks. No change to *which* tests run — only how they're grouped for the parallel run. Estimated ~4-5 min faster per provider DB cell on the existing free runner (no bigger runner / no extra DB containers). --- ##### Was generative AI tooling used to co-author this PR? - [X] Yes — Claude Code (Opus 4.8) Generated-by: Claude Code (Opus 4.8) following [the guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
