DNS Group Metrics¶

The dns-operator exposes Prometheus metrics for monitoring DNS group assignments and failover operations. These metrics are available when the dns-operator is running and groups are configured.

Available Metrics¶

dns_record_group_info¶

Field	Value
Type	Gauge
Description	Reports the group assignment for each DNSRecord. Always emits a value of 1; the group is identified by the `group` label.
Labels	`dns_record_name`, `dns_record_namespace`, `group`

dns_record_group_active¶

Field	Value
Type	Gauge
Description	Reports whether each DNSRecord is in an active group (1) or inactive group (0). Based on the `Active` status condition of the DNSRecord. Also emitted for ungrouped records (with an empty `group` label) — use `{group!=""}` to filter these out.
Labels	`dns_record_name`, `dns_record_namespace`, `group`

dns_record_inactive_group_cleanup_total¶

Field	Value
Type	Counter
Description	Counts cleanup operations that remove DNS records or TXT registry entries belonging to inactive groups. Incremented each time an active controller cleans up records from an inactive group.
Labels	`dns_record_name`, `dns_record_namespace`, `group`

Example Queries¶

List all DNSRecords and their group assignments¶

dns_record_group_info

Show only inactive DNSRecords¶

dns_record_group_active{group!=""} == 0

Show only active DNSRecords¶

dns_record_group_active{group!=""} == 1

Count DNSRecords per group¶

count by (group) (dns_record_group_info)

Count active records per group¶

count by (group) (dns_record_group_active{group!=""} == 1)

Count inactive records per group¶

count by (group) (dns_record_group_active{group!=""} == 0)

Rate of inactive group cleanup operations (per minute)¶

rate(dns_record_inactive_group_cleanup_total[5m]) * 60

Alert: inactive group cleanup is occurring¶

This can indicate that a failover has been triggered and active controllers are cleaning up records from the now-inactive group.

increase(dns_record_inactive_group_cleanup_total[10m]) > 0

Alert: unexpected inactive group cleanup activity¶

Sustained cleanup activity may indicate that a group was deactivated unintentionally or that an operator is stuck cycling between active and inactive states:

apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata:
  name: dns-group-alerts
spec:
  groups:

    - name: dns-groups
      rules:
        - alert: DNSInactiveGroupCleanupSustained
          expr: rate(dns_record_inactive_group_cleanup_total[5m]) > 0
          for: 30m
          labels:
            severity: warning
          annotations:
            summary: "Sustained inactive group cleanup for {{ $labels.dns_record_name }} in group {{ $labels.group }}"
            description: "Continuous cleanup operations for an inactive group over 30 minutes may indicate a misconfigured group or an operator cycling between states."

Note: Records in a standby (inactive) group are expected — dns_record_group_active == 0 alone is not an error condition. Alert on cleanup activity or unexpected group transitions rather than on inactive state directly.

Grafana Dashboard Tips¶

Use a stat panel with count by (group) (dns_record_group_info) to show the number of records per group.
Use a table panel with dns_record_group_info * on(dns_record_name, dns_record_namespace) group_left dns_record_group_active to show each record's group and active state together.
Use a time series panel with rate(dns_record_inactive_group_cleanup_total[5m]) to visualize cleanup activity over time, which correlates with failover events.