monitoring-ng/alerts/node.yml.tmpl

apiVersion: 1

# List of alert rule UIDs that should be deleted
deleteRules:
  {{ if not (env "ALERT_NODE_DISK_SPACE_LEFT") }}
  - orgId: 1
    uid: coopcloud_node_disk_space_left
  {{ end }}
  {{ if not (env "ALERT_NODE_MEMORY_USAGE") }}
  - orgId: 1
    uid: coopcloud_node_memory_usage
  {{ end }}

groups:
- orgId: 1
  name: node
  folder: node
  interval: 5m
  rules:
    {{ if (env "ALERT_NODE_DISK_SPACE_LEFT") }}
    - uid: coopcloud_node_disk_space_left
      title: Node Disk Space
      condition: C
      data:
        - refId: A
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: PBFA97CFB590B2093
          model:
            editorMode: code
            expr: (node_filesystem_free_bytes{fstype="ext4"} / node_filesystem_size_bytes{fstype="ext4"}) * 100
            instant: true
            intervalMs: 1000
            legendFormat: __auto
            maxDataPoints: 43200
            range: false
            refId: A
        - refId: C
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: __expr__
          model:
            conditions:
                - evaluator:
                    params:
                        - {{ env "ALERT_NODE_DISK_SPACE_LEFT" }}
                    type: lt
                  operator:
                    type: and
                  query:
                    params:
                        - C
                  reducer:
                    params: []
                    type: last
                  type: query
            datasource:
                type: __expr__
                uid: __expr__
            expression: A
            intervalMs: 1000
            maxDataPoints: 43200
            refId: C
            type: threshold
      noDataState: NoData
      execErrState: Error
      for: 5m
      annotations:
        description: ""
        runbook_url: ""
        summary: Less than {{ env "ALERT_NODE_DISK_SPACE_LEFT" }}% disk space left on {{`{{ $labels.instance }}`}} ({{`{{ (index $values "A").Value }}`}}% left)
      labels:
        "": ""
      isPaused: false
    {{ end }}
    {{ if (env "ALERT_NODE_MEMORY_USAGE")  }}
    - uid: coopcloud_node_memory_usage
      title: Node Memory Usage
      condition: C
      data:
        - refId: A
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: PBFA97CFB590B2093
          model:
            editorMode: code
            expr: (node_memory_MemAvailable_bytes / node_memory_MemTotal_bytes) * 100
            instant: true
            intervalMs: 1000
            legendFormat: __auto
            maxDataPoints: 43200
            range: false
            refId: A
        - refId: C
          relativeTimeRange:
            from: 600
            to: 0
          datasourceUid: __expr__
          model:
            conditions:
                - evaluator:
                    params:
                        - {{ env "ALERT_NODE_MEMORY_USAGE"  }}
                    type: gt
                  operator:
                    type: and
                  query:
                    params:
                        - C
                  reducer:
                    params: []
                    type: last
                  type: query
            datasource:
                type: __expr__
                uid: __expr__
            expression: A
            intervalMs: 1000
            maxDataPoints: 43200
            refId: C
            type: threshold
      noDataState: NoData
      execErrState: Error
      for: 5m
      annotations:
        summary: Memory usage is above {{ env "ALERT_NODE_MEMORY_USAGE"  }}% on {{`{{ $labels.instance }}`}}  ({{`{{ printf "%.2f" (index $values "A").Value }}`}}%  usage)
      isPaused: false
    {{ end }}