We're using the Dragonfly operator to provision a 3-replica Dragonfly instance in our cluster. The operator is expected to elect a master pod and apply the role=master label so that the Dragonfly service can select the master pod via this label. However, the label is not being applied, and the Dragonfly service remains without endpoints, causing downstream issues.
Logs:
Logs from the Dragonfly operator show repeated messages like:
INFO Received {"controller": "pod", "controllerGroup": "", "controllerKind": "Pod", "Pod": {"name":"<name>-dragonfly-0","namespace":"<name>"}}
INFO Dragonfly object is not initialized yet {"controller": "pod", "controllerGroup": "", "controllerKind": "Pod"}
Context:
Upon further inspection, I noticed that:
The status field is missing entirely from the Dragonfly CR, which aligns with the logs saying the object is not being fully initialized.
The associated PVCs for the Dragonfly pods are correctly bound, so storage does not appear to be the issue.
The Dragonfly resources were restored via Velero, so I suspect the issue may have to do with restoration side-effects, though I haven't been able to find any specific issues.
I have tried destroying and recreating all related resources (i.e., the operator, the Dragonfly CR, etc.) just to rule out any anomalies, and the issue persists.
Summary:
We're using the Dragonfly operator to provision a 3-replica Dragonfly instance in our cluster. The operator is expected to elect a master pod and apply the
role=masterlabel so that the Dragonfly service can select the master pod via this label. However, the label is not being applied, and the Dragonfly service remains without endpoints, causing downstream issues.Logs:
Logs from the Dragonfly operator show repeated messages like:
Context:
Upon further inspection, I noticed that:
statusfield is missing entirely from the Dragonfly CR, which aligns with the logs saying the object is not being fully initialized.I have tried destroying and recreating all related resources (i.e., the operator, the Dragonfly CR, etc.) just to rule out any anomalies, and the issue persists.
Manifests:
Dragonfly
```yaml --- apiVersion: dragonflydb.io/v1alpha1 kind: Dragonfly metadata: labels: app.kubernetes.io/name: dragonfly name:Objects:
Dragonfly (note: no status field)
```yaml --- apiVersion: dragonflydb.io/v1alpha1 kind: Dragonfly metadata: annotations: kubectl.kubernetes.io/last-applied-configuration:StatefulSet
```yaml --- apiVersion: apps/v1 kind: StatefulSet metadata: name:Pod (1 of 3)
```yaml --- apiVersion: v1 kind: Pod metadata: name:PVC (1 of 3)
```yaml --- apiVersion: v1 kind: PersistentVolumeClaim metadata: name: df-PV (1 of 3)
```yaml --- apiVersion: v1 kind: PersistentVolume metadata: name: pvc-Service
```yaml --- apiVersion: v1 kind: Service metadata: name:Endpoints (note: no data)
```yaml --- apiVersion: v1 kind: Endpoints metadata: name:Environment:
Please let me know if you need more information or logs. Thanks!