[flaky test] TopologyAwareScheduling when Scaling a Job requesting TAS via workload slices should preserve topology assignment during scale-up

[mimowo]

Which test is flaking?:

TopologyAwareScheduling when Scaling a Job requesting TAS via workload slices should preserve topology assignment during scale-up

Link to failed CI job or steps to reproduce locally:

https://prow.k8s.io/view/gs/kubernetes-ci-logs/pr-logs/pull/kubernetes-sigs_kueue/9224/pull-kueue-test-e2e-main-1-35/2023720289359106048

Failure message or logs:

{Expected
    <*v1beta2.Workload | 0x0>: nil
not to be nil failed [FAILED] Expected
    <*v1beta2.Workload | 0x0>: nil
not to be nil
In [It] at: /home/prow/go/src/sigs.k8s.io/kueue/test/e2e/singlecluster/tas_test.go:643 @ 02/17/26 11:39:16.134
}

Anything else we need to know?:

recently introduced in #8580

[mimowo]

/assign @sohankunkerkar
Tentatively, who works on the TAS+Elastic Workloads.

[mimowo]

https://prow.k8s.io/view/gs/kubernetes-ci-logs/pr-logs/pull/kubernetes-sigs_kueue/9311/pull-kueue-test-e2e-main-1-35/2023724461684953088

[mimowo]

This seems to be too common, let me disable temporarily: #9317