damccorm opened a new pull request, #41125: URL: https://github.com/apache/beam/pull/41125
### Summary In `IcebergIO Integration Tests` (e.g., https://github.com/apache/beam/actions/runs/42499259877), `HadoopCatalogIT` (23 tests) and `HadoopCatalogCdcWriteIT` (8 tests) failed within ~1 second during `@Before setUp()` when GKE Metadata Server (`metadata.google.internal`) returned a transient `503 Service Unavailable` during `GoogleHadoopFileSystem` initial OAuth token fetch. Because `setUp()` immediately fails without retry/backoff, JUnit burned through all 31 test methods during the 1-second metadata server blip. This change: - Retries `createCatalog()` and namespace creation up to 3 times with backoff in `IcebergCatalogBaseIT.setUp()` and `IcebergCdcWriteBaseIT.setUp()`. - Only registers `namespace()` in `namespacesToCleanup` once setup succeeds so `@After cleanUp()` does not fail with secondary exceptions if setup fails. ### Verification - Triggered `IcebergIO Integration Tests` on branch `fix/iceberg-catalog-it-setup-retry`: https://github.com/apache/beam/actions/runs/42660534685 -- 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]
