GoogleCloudDataproc
[initialization-actions] Please add support for http proxy along with tests which exercise the use case
Some of our more security-conscious customers depend on http proxies to fetch installation artifacts.
Customers have requested that we ensure that initialization actions work correctly and perform well when executed in environments which depend on HTTP proxies for access to the internet.
I've had another user request better support for environments with HTTP proxies.
Hello everyone, thank you for your patience and continued interest in this critical feature.
I'm happy to provide an update: we are actively making progress on adding robust HTTP proxy support to our initialization actions.
To ensure we address this correctly and prevent future regressions for this and other complex networking scenarios, our first step has been to build out a comprehensive, end-to-end testing framework. This framework allows us to automatically provision complex network environments, including a dedicated proxy server, and then launch Dataproc clusters within that environment to validate that proxy settings are respected throughout the entire cluster lifecycle.
We have now completed the design and initial draft of this test framework and the specific test case for the proxy-egress scenario. Our next step is to exercise this test suite to validate the code changes required in the initialization actions themselves.
We are on track to have this work completed, tested, and released during Q3 2025. We appreciate how important this functionality is for running Dataproc in enterprise environments and will post further updates here as we approach the release.