test: skip virError: 'internal error: client socket is closed'

[universal-itengineer]

Description

Intermittent KubeVirt migration failures with virError(Code=1, Domain=7, Message='internal error: client socket is closed') make volume migration e2e scenarios flaky and produce false-negative test results.
This change introduces a temporary, explicit skip path for this known external issue in StorageClassMigration and RWOVirtualDiskMigration flows.

Why do we need it, and what problem does it solve?

• Stabilizes e2e pipelines by preventing failures caused by a known upstream/libvirt socket-closure issue rather than product logic.
• Applies skip handling in all relevant migration wait paths (shared helpers and local Eventually checks) so flaky failures are caught consistently.
• Reads migration failure reason from intvirtvmi.Status.MigrationState.FailureReason and matches robustly (tolerant to YAML formatting differences such as quotes/newlines).
• Adds clear TODO-marked temporary behavior to remove once kubevirt-side issue is fixed.

What is the expected result?

• Tests in StorageClassMigration and RWODiskMigration no longer fail on the known client socket is closed migration error.
• Affected scenarios are marked as Skipped with explicit context instead of reported as product regressions.
• Normal migration success/failure behavior for other reasons remains unchanged.

Checklist

• The code is covered by unit tests.
• e2e tests passed.
• Documentation updated according to the changes.
• Changes were tested in the Kubernetes cluster manually.

Changelog entries