kubernetes/kops · error

error describing autoScalingGroups: %v

Error message

error describing autoScalingGroups: %v

What it means

deregisterInstance() re-fetches the Auto Scaling Group via DescribeAutoScalingGroups to learn which Classic ELBs and target groups the instance must be drained from; if that AWS API call fails, this error wraps the SDK error. Without the ASG details kOps cannot determine the load balancer set, so deregistration aborts.

Source

Thrown at upup/pkg/fi/cloudup/awsup/aws_cloud.go:522

		} else {
			return fmt.Errorf("error deleting instance %q: %v", id, err)
		}
	}

	klog.V(8).Infof("deleted aws ec2 instance %q", id)

	return nil
}

// deregisterInstance ensures that the instance is fully drained/removed from all associated loadBalancers and targetGroups before termination.
func deregisterInstance(ctx context.Context, c AWSCloud, i *cloudinstances.CloudInstance) error {
	asg := i.CloudInstanceGroup.Raw.(*autoscalingtypes.AutoScalingGroup)

	asgDetails, err := c.Autoscaling().DescribeAutoScalingGroups(ctx, &autoscaling.DescribeAutoScalingGroupsInput{
		AutoScalingGroupNames: []string{aws.ToString(asg.AutoScalingGroupName)},
	})
	if err != nil {
		return fmt.Errorf("error describing autoScalingGroups: %v", err)
	}

	if len(asgDetails.AutoScalingGroups) == 0 {
		return nil
	}

	// there will always be only one ASG in the DescribeAutoScalingGroups response.
	loadBalancerNames := asgDetails.AutoScalingGroups[0].LoadBalancerNames
	targetGroupArns := asgDetails.AutoScalingGroups[0].TargetGroupARNs

	eg, _ := errgroup.WithContext(context.Background())

	if len(loadBalancerNames) != 0 {
		eg.Go(func() error {
			return deregisterInstanceFromClassicLoadBalancer(ctx, c, loadBalancerNames, i.ID)
		})
	}

View on GitHub (pinned to 4c8573c808)

Solutions

1. Verify the IAM policy grants autoscaling:DescribeAutoScalingGroups
2. Check the ASG still exists (name comes from the cached CloudInstanceGroup.Raw) — if deleted, refresh cluster state and retry
3. Retry the operation; throttling errors (Throttling/RequestLimitExceeded) are transient — reduce concurrency for huge updates
4. Check AWS region configuration matches where the cluster's ASGs live

Example fix

// before: policy missing
// error describing autoScalingGroups: AccessDenied: User is not authorized to perform: autoscaling:DescribeAutoScalingGroups
// after: add to IAM policy statement
{"Effect":"Allow","Action":["autoscaling:DescribeAutoScalingGroups"],"Resource":"*"}

_, err := asgSvc.DescribeAutoScalingGroups(&autoscaling.DescribeAutoScalingGroupsInput{
	AutoScalingGroupNames: []string{asgName}})
if err != nil {
	return fmt.Errorf("precheck failed: %w", err)
}

func isThrottle(err error) bool {
	var ae smithy.APIError
	return errors.As(err, &ae) && (ae.ErrorCode() == "Throttling" || ae.ErrorCode() == "RequestLimitExceeded")
}

if err := cloud.DeregisterInstance(inst); err != nil {
	if strings.Contains(err.Error(), "error describing autoScalingGroups") {
		// retry with backoff; throttling/describe failures are usually transient
		return retryWithBackoff(3, 5*time.Second, func() error { return cloud.DeregisterInstance(inst) })
	}
	return err
}

When it happens

Trigger: deregisterInstance calls c.Autoscaling().DescribeAutoScalingGroups with the instance's ASG name and AWS returns an error: throttling (RequestLimitExceeded), AccessDenied, InvalidGroupName.NotFound, or network failure.

Common situations: IAM policy missing autoscaling:DescribeAutoScalingGroups; ASG deleted while kOps still holds a reference to it; API throttling during large rolling updates across many groups; temporary AWS API outage.

Related errors

• error deleting autoscaling group %q: %v
• error listing AutoScalingGroups: %v
• error creating AutoScalingGroup: %s
• error listing ASG Lifecycle Hooks: %v
• DIGITALOCEAN_ACCESS_TOKEN is required

AI-assisted analysis of kubernetes/kops@4c8573c808 (2026-09-05). Data as JSON: /api/errors/967aa017559b3edd. Report an issue: GitHub.