What causes "Invalid value for %v: %v" in kubernetes/kops?

Set BOOT_VOLUME_SIZE metadata to a plain base-10 integer of gigabytes, e.g. "50" instead of "50GB"

Invalid value for %v: %v

Error message

Invalid value for %v: %v

What it means

Thrown by includeBootVolumeOptions when the metadata key openstack.BOOT_VOLUME_SIZE (kops.openstack.org/boot-volume-size) cannot be parsed by strconv.ParseInt(s, 10, 64). This is a pure client-side validation error before any Nova call — the metadata value must be a base-10 integer representing the boot volume size in GB. The wrapped strconv.NumError is included in the message.

Source

Thrown at upup/pkg/fi/cloudup/openstacktasks/instance.go:447

	i, err := t.Cloud.GetImage(fi.ValueOf(e.Image))
	if err != nil {
		return servers.CreateOpts{}, fmt.Errorf("Error getting image information: %v", err)
	}

	blockDevice := servers.BlockDevice{
		BootIndex:           0,
		DeleteOnTermination: true,
		DestinationType:     "volume",
		SourceType:          "image",
		UUID:                i.ID,
		VolumeSize:          i.MinDiskGigabytes,
	}

	if s, ok := e.Metadata[openstack.BOOT_VOLUME_SIZE]; ok {
		i, err := strconv.ParseInt(s, 10, 64)
		if err != nil {
			return servers.CreateOpts{}, fmt.Errorf("Invalid value for %v: %v", openstack.BOOT_VOLUME_SIZE, err)
		}

		blockDevice.VolumeSize = int(i)
	}

	opts.BlockDevice = []servers.BlockDevice{blockDevice}
	return opts, nil
}

func bootFromVolume(m map[string]string) bool {
	v, ok := m[openstack.BOOT_FROM_VOLUME]
	if !ok {
		return false
	}

	switch v {
	case "true", "enabled":
		return true

View on GitHub (pinned to 4c8573c808)

Solutions

Set BOOT_VOLUME_SIZE metadata to a plain base-10 integer of gigabytes, e.g. "50" instead of "50GB"
Verify what was actually set: inspect the instance task metadata in the cluster spec / `kops toolbox template` output and fix the value
If a unit is desired, convert manually (50GB → 50) — the code does no unit parsing
Rerun `kops update cluster` after correcting; also fix any automation (scripts/Helm values) generating the metadata

Example fix

// before (cluster spec metadata)
kops.k8s.io/boot-volume-size: "100GB"
// after
kops.k8s.io/boot-volume-size: "100"

Defensive patterns

Strategy: validation

Validate before calling

// validate BOOT_VOLUME_SIZE metadata is a plain base-10 integer before apply
func validateBootVolumeSize(metadata map[string]string) error {
	s, ok := metadata[openstack.BOOT_VOLUME_SIZE]
	if !ok {
		return nil
	}
	if _, err := strconv.ParseInt(s, 10, 64); err != nil {
		return fmt.Errorf("%s must be a plain integer of GB, got %q: %w", openstack.BOOT_VOLUME_SIZE, s, err)
	}
	return nil
}

Type guard

func isBootVolumeSizeValid(metadata map[string]string) bool {
	s, ok := metadata[openstack.BOOT_VOLUME_SIZE]
	if !ok {
		return true
	}
	_, err := strconv.ParseInt(s, 10, 64)
	return err == nil
}

Prevention

Always specify boot volume size as bare GB integers ("100"), never "100GB" or "100 GiB"
Lint cluster templates to check the metadata value parses as an int before apply
Avoid shell/templated values that may render empty strings into metadata
Ensure generated specs don't carry decimal or thousands-separated numbers
Document the expected format where team members copy cluster specs

When it happens

Trigger: Instance metadata contains a non-numeric boot-volume-size value, e.g. "50GB", "50Gi", "0x32", "50.0", or an empty string; set via cluster spec openstack metadata or blockStore config. Parse fails with `strconv.ParseInt: parsing "...": invalid syntax` (or value out of range for int64).

Common situations: User wrote "50GB" or "50 GiB" instead of plain "50"; value copied from Terraform-style config that allows units; shell interpolation produced an empty string; decimal size like "12.5"; value set programmatically as a formatted number with thousands separator.

Related errors

AI-assisted analysis of kubernetes/kops@4c8573c808 (2026-09-05). Data as JSON: /api/errors/8351401674bfa27e. Report an issue: GitHub.