Question

A stateful application uses a PVC bound to a PV backed by a block-storage CSI driver that only supports ReadWriteOnce. The PVC is currently mounted by a pod running on node worker-1. You scale the Deployment to 2 replicas, and the second pod is scheduled to worker-2 but remains stuck in ContainerCreating. Events on the second pod show a FailedAttachVolume / Multi-Attach error. What is the correct explanation and remedy?

Accepted Answer

ReadWriteOnce allows the volume to be mounted read-write by only one node at a time; use a ReadWriteMany-capable storage backend or ensure both pods land on the same node.

Answer

ReadWriteOnce means only one pod can ever use the volume; delete and recreate the PVC with ReadWriteOncePod to allow the second pod to attach.

Answer

The PV reclaim policy is set to Retain, which blocks additional attachments; change it to Delete so the volume can be shared across nodes.

Answer

The volumeBindingMode is Immediate, preventing multi-node attachment; recreate the StorageClass with WaitForFirstConsumer to allow simultaneous mounts.