Run portable ValidatesRunner tests one at a time

[aIbrahiim]

fixes: #38934
caused by: #40324
#38934 (comment)

[github-actions]

📋 Review Summary

This PR limits in-JVM test concurrency for Flink Portable ValidatesRunner tasks to 1 thread (beam.validatesRunner.parallelThreads=1), matching the existing configuration in the Spark job server. This resolves OOMs and test hangs caused by spawning up to 8 concurrent in-process job servers and Flink mini-clusters within a single 4GB test worker JVM.

🔍 General Feedback

• The fix is clean, well-targeted, and directly addresses the resource contention introduced when beam.test.parallelThreads was defaulted to 8 in #40324 while preserving fork-level parallelism (maxParallelForks = 2).

---

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

[Abacn]

Thanks!

[Abacn]

However it seems the test still stuck, probably other issues introduced since then

[github-actions]

Checks are failing. Will not request review until checks are succeeding. If you'd like to override that behavior, comment assign set of reviewers

[aIbrahiim]

However it seems the test still stuck, probably other issues introduced since then

yeah the 1.20 job is the old DataSet suite and those 7 tests already hang on master until they hit the 20 min timeout but the other Flink versions passed. i excluded them for DataSet only so that job can actually finish

[Abacn]

Thanks, lgtm for now, however we still need to fix these tests as Flink 1.20 is LTS. Could you please open an Issue as a release block?

[Abacn]

actually it seems still stuck