groups:
  - name: dedicated-regionalserver-nodepool-gc
    rules:
      - alert: NodePoolGCFailed
        annotations:
          summary: BR nodepool garbage collection failed
          description: >-
            BR nodepool {{ $labels.namespace }}/{{ $labels.name }} is in the
            failed GC state (provider={{ $labels.provider }},
            region={{ $labels.region }}, component={{ $labels.component }},
            tenant_id={{ $labels.tenant_id }},
            project_id={{ $labels.project_id }},
            cluster_id={{ $labels.cluster_id }},
            reason={{ $labels.reason }}). Deletion is blocked until the
            underlying safety check or GC operation failure is resolved.
          runbook_url: https://github.com/tidbcloud/runbooks/wiki/NodePoolGCFailed
        expr: |
          infra_br_nodepool_gc_status{service="regional-server"} == 4
        for: 5m
        labels:
          severity: major
          response_class: ticket
          component: br
          oncall_team: tidb_x_tools
