Fix: Kubernetes PVC Stuck in Pending State

!Symptoms

• PVC status shows Pending indefinitely
• Pods using the PVC are stuck in Pending state
• kubectl describe pvc shows 'waiting for first consumer'
• No PersistentVolume is being dynamically provisioned

?Root Causes

• No StorageClass configured or default StorageClass missing
• StorageClass provisioner not installed (e.g., EBS CSI driver, NFS provisioner)
• WaitForFirstConsumer binding mode but no pod scheduled to the same zone
• Insufficient cloud provider quota for volumes (EBS, Persistent Disk)
• Storage class parameters don't match available zones or disk types

#Diagnosis Steps

1. 1Run `kubectl describe pvc <name>` and check events for provisioning errors
2. 2Run `kubectl get sc` to verify StorageClass exists and check the default annotation
3. 3Run `kubectl get pv` to see if a matching PersistentVolume exists
4. 4Check cloud provider quotas: `aws service-quotas` or GCP console
5. 5Verify CSI driver pods are running: `kubectl get pods -n kube-system | grep csi`

>Fix

1. 1Install the required CSI driver (e.g., `aws-ebs-csi-driver`, `gcp-pd-csi-driver`)
2. 2Create a default StorageClass: add `storageclass.kubernetes.io/is-default-class: true` annotation
3. 3If WaitForFirstConsumer, ensure a pod referencing the PVC is scheduled
4. 4Increase cloud provider volume quotas if at limit
5. 5Manually create a PV matching the PVC's requirements as a temporary fix

*Prevention

• Include CSI drivers in cluster bootstrapping automation
• Always define a default StorageClass in your cluster config
• Monitor cloud provider quotas with alerts
• Use Immediate binding mode for non-topology-constrained workloads
• Test PVC provisioning in staging before relying on it in production

Related Error Messages