[shunping]

GcsUtil now uses GcsUtilV2 (java-storage) by default instead of GcsUtilV1 (gcsio). Before this, V2 was opt-in through the use_gcsutil_v2 experiment.

Experiments

Experiment	Behavior
(none)	GcsUtil V2 (new default)
disable_gcsutil_v2	New. Falls back to GcsUtil V1
use_grpc_for_gcs	Falls back to GcsUtil V1, since only gcsio supports gRPC
use_gcsutil_v2	Accepted but does nothing now. Kept so existing pipelines don't break

Either fallback experiment wins over use_gcsutil_v2. The names follow the existing use_runner_v2 / disable_runner_v2 pair.

[shunping]

r: @damccorm

[github-actions]

Stopping reviewer notifications for this pull request: review requested by someone other than the bot, ceding control. If you'd like to restart, comment assign set of reviewers

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.04%. Comparing base (a8d6afd) to head (b4eb547).
⚠️ Report is 35 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #40713      +/-   ##
==========================================
- Coverage   80.04%   80.04%   -0.01%     
==========================================
  Files         490      490              
  Lines       82916    82888      -28     
==========================================
- Hits        66373    66350      -23     
+ Misses      16543    16538       -5     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

📋 Review Summary

This PR switches GcsUtil to use GcsUtilV2 (java-storage) by default while providing --experiments=disable_gcsutil_v2 (and preserving use_grpc_for_gcs) to fall back to GcsUtilV1 (gcsio). It also fixes toStorageObject(Blob) to preserve kmsKeyName when delegating getObject to V2 and updates the corresponding unit, integration, and load tests.

🔍 General Feedback

• The implementation and fallback precedence in GcsUtil are clean, well-documented, and thoroughly covered by unit and integration tests.
• Good catch preserving kmsKeyName in GcsUtil.toStorageObject(Blob) and propagating the experiment in GcsUtilIT.testRewriteMultiPart.
• One minor note: the two newly added .github/trigger_files/beam_PostCommit_Java_Examples_*.json files don't appear to have corresponding workflows listening to them in .github/workflows/ (unlike beam_PostCommit_Java_Examples_Dataflow_Java.json).

---

💡 Tip: Comment /gemini review on this pull request to trigger a new Gemini code review.

[damccorm]

Do we plan to enable grpc with v2 ever? Or why is this scoped out?

[shunping]

It is in our plan to support gRPC in V2. However, since gRPC support in V1 is still controlled by an experiment, we didn't want it to block the default mode (HTTP support) switch to V2.

In addition, we'll need to run another round of benchmarks comparing HTTP/gRPC modes on V2 and gRPC mode on V1/V2 to make sure there are no discrepancies before we route gRPC support to V2.

[damccorm]

SGTM

[damccorm]

Thanks!