fix(pool): guard decommission lifecycle recovery

[GatewayJ]

Type of Change

• New Feature
• Bug Fix
• Documentation
• Performance Improvement
• Test/CI
• Refactor
• Other:

Related Issues

Fixes rustfs/backlog#1078
Fixes rustfs/backlog#1079
Related to rustfs/backlog#1070

Summary of Changes

• Count only live pools when evaluating pool decommission start requests, so pools already decommissioning or decommissioned cannot keep the last remaining live pool eligible for decommission.
• Reject a batch of live-pool Start requests before issuing RustFS admin calls when that batch would decommission every live pool in the Tenant.
• Restore normal workload reconciliation after a canceled or failed decommission request is removed, instead of leaving the pool permanently lifecycle-gated.
• Add focused regression tests for last-live-pool protection, batch start protection, and terminal lifecycle recovery.

Checklist

• I have read and followed the CONTRIBUTING.md guidelines
• Passed make pre-commit (fmt-check + clippy + test + console-lint + console-fmt-check)
• Added/updated necessary tests
• Documentation updated (if needed)
• CHANGELOG.md updated under [Unreleased] (if user-visible change)
• CI/CD passed (if applicable)

Impact

• Breaking change (CRD/API compatibility)
• Requires doc/config/deployment update
• Other impact: Preserves normal StatefulSet reconciliation after canceled or failed decommission requests are cleared; prevents decommissioning all live pools, including multi-request batches.

Verification

make pre-commit

Additional Notes

This change only adjusts operator lifecycle gating. It does not change RustFS admin API calls, CRD schema, StatefulSet deletion behavior, or PVC retention behavior.