damccorm opened a new pull request, #41167: URL: https://github.com/apache/beam/pull/41167
### Summary Fixes intermittent `TimeoutException` failures when starting SDK harness containers via `DockerCommand.runImage` in `PostCommit Go VR Spark` ([run 42749923028](https://github.com/apache/beam/actions/runs/42749923028)) and `PostCommit Go VR Flink` ([run 42750055749](https://github.com/apache/beam/actions/runs/42750055749)). ### Root Cause `DockerCommand.runImage` previously executed `docker pull <imageTag>` unconditionally on every SDK harness container startup (including locally built dev images like `apache/beam_go_sdk:2.79.0.dev`), adding unnecessary remote registry requests and Docker daemon contention across every test in `ValidatesRunner` suites, and did not retry `docker run -d` if the daemon timed out once (e.g. while Testcontainers simultaneously started/stopped test containers in `TestMongoDBIO_Write` / `TestMongoDBIO_Read`). ### Fix 1. Check whether `imageTag` is already present locally via `docker image inspect <imageTag>` before attempting `docker pull <imageTag>`. 2. Retry `docker run -d` once if it fails with `TimeoutException`. ### Verification - Triggered `beam_PostCommit_Go_VR_Spark.yml` workflow run on branch: [`42841205302`](https://github.com/apache/beam/actions/runs/42841205302) -- 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]
