potiuk opened a new pull request, #71544: URL: https://github.com/apache/airflow/pull/71544
The lowest-dependency providers job fails when IBM's download server stops answering: the `ibm.mq` pre-extras manifest fetch passed no socket timeout, so each of the five routes (DNS plus the four published anycast IPs) sat in the kernel's TCP connect timeout, and the single pass over them had no retry. One transient upstream outage took the whole job down — see the `Errno 110` failures in [this run](https://github.com/apache/airflow/actions/runs/31479252918/job/93745832553). Every attempt now uses a 20s socket timeout, and the full set of routes is retried in 3 rounds with a short sleep between them, so a stalled transfer or a blackholed IP recovers in seconds. A checksum mismatch is still fatal on the first attempt — that means the manifest disagrees with what upstream serves, which retrying cannot fix. --- ##### Was generative AI tooling used to co-author this PR? - [X] Yes — Claude Code (Opus 5) Generated-by: Claude Code (Opus 5) 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]
