How to Recycle a Node Pool Instance in a Vultr Kubernetes Engine Cluster

1 min readUpdated 4 November 2026

Replace one node in a Vultr Kubernetes Engine node pool with a fresh instance. The old node's local data is lost.

Recycling a node pool instance in a Vultr Kubernetes Engine (VKE) cluster replaces a specific node with a fresh instance of the same configuration. Recycle a node to resolve node-level issues, apply updates, or recover from resource constraints without disrupting the rest of the node pool. VKE provisions the replacement node and waits for it to join the cluster before it drains and removes the target node, so the node pool temporarily runs one node above its configured size. The Vultr Console labels this action Replace Node.

Warning

Recycling a node deletes the target instance, including any ephemeral data, logs, or local configurations on it, after the replacement joins the cluster. Draining reschedules workloads onto the replacement node or other available nodes. Linked resources remain attached, but if the node is part of a Load Balancer backend, brief downtime can occur while the new node registers. To prevent data loss, store important data in PersistentVolumes or external storage, or back it up before recycling.

Follow this guide to recycle a node in a Vultr Kubernetes Engine cluster using the Vultr Console, API, or CLI.

Navigate to Compute and click Kubernetes.
Click the target cluster to open its management page.
Click the Nodes tab.
Click + next to the target node pool name to view its nodes.
Click the Replace Node icon in the row of the target node.
In the Replace Node? dialog, verify that the Node field shows the correct node name.
Select Yes, replace this node. and click Replace Node to recycle the node.

0 Comments

Start the conversation

Be the first to comment and share your perspective with the community.

Recycle a Node Pool Instance in a Vultr Kubernetes Cluster | Vultr Docs