kubernetes/kops · error

error detaching instance %q: %v

Error message

error detaching instance %q: %v

What it means

Same detach failure as its sibling error, but for instances with no associated Kubernetes node (nodeName == ""), e.g. control-plane or bastion instances identified only by cloud ID. It wraps the cloud provider DetachInstance error with just the instance ID.

Source

Thrown at pkg/instancegroups/instancegroups.go:640

// detachInstance detaches a Cloud Instance
func (c *RollingUpdateCluster) detachInstance(u *cloudinstances.CloudInstance) error {
	id := u.ID
	nodeName := ""
	if u.Node != nil {
		nodeName = u.Node.Name
	}
	if nodeName != "" {
		klog.Infof("Detaching instance %q, node %q, in group %q.", id, nodeName, u.CloudInstanceGroup.HumanName)
	} else {
		klog.Infof("Detaching instance %q, in group %q.", id, u.CloudInstanceGroup.HumanName)
	}

	if err := c.Cloud.DetachInstance(u); err != nil {
		if nodeName != "" {
			return fmt.Errorf("error detaching instance %q, node %q: %v", id, nodeName, err)
		}
		return fmt.Errorf("error detaching instance %q: %v", id, err)
	}

	return nil
}

// deleteInstance deletes an Cloud Instance.
func (c *RollingUpdateCluster) deleteInstance(u *cloudinstances.CloudInstance) error {
	id := u.ID
	nodeName := ""
	if u.Node != nil {
		nodeName = u.Node.Name
	}
	if nodeName != "" {
		klog.Infof("Stopping instance %q, node %q, in group %q (this may take a while).", id, nodeName, u.CloudInstanceGroup.HumanName)
	} else {
		klog.Infof("Stopping instance %q, in group %q (this may take a while).", id, u.CloudInstanceGroup.HumanName)
	}

View on GitHub (pinned to 4c8573c808)

Solutions

1. Re-run rolling update to refresh group membership
2. Verify IAM permissions for detaching instances
3. Check the instance still exists and belongs to the group in the cloud console
4. Retry after transient cloud API errors

// verify the instance exists in the cloud provider before detaching
inst, err := ec2Client.DescribeInstances(&ec2.DescribeInstancesInput{InstanceIds: []*string{&u.ID}})
// if not found or not in the expected group, refresh group state first

func detachable(u *cloudinstances.CloudInstance) bool { return u != nil && u.ID != "" }

if err := c.Cloud.DetachInstance(u); err != nil {
    if isNotFound(err) { return nil } // instance already gone
    return fmt.Errorf("error detaching instance %q: %v", id, err)
}

When it happens

Trigger: c.Cloud.DetachInstance(u) fails and nodeName == "": detach of a control-plane/bastion instance rejected by the cloud API — instance not in group, permission denied, throttling, or instance already terminated.

Common situations: Rolling the control-plane of a cluster whose instances were modified out-of-band; IAM misconfiguration; transient cloud API errors during etcd member replacement rolls.

Related errors

• error detaching instance %q, node %q: %v
• error deleting instance %q, node %q: %v
• error deleting instance %q: %v
• failed to detach instance: %v
• spotinst does not support surging

AI-assisted analysis of kubernetes/kops@4c8573c808 (2026-09-05). Data as JSON: /api/errors/b88e02948bb873a5. Report an issue: GitHub.