kubernetes/kops · error

reading from stdin: %v

Error message

reading from stdin: %v

What it means

Returned by `kops delete -f -` when `ConsumeStdin` fails while reading the manifests from standard input. kOps reads the full stdin stream as the resource spec(s) to delete; an error reading the stream (e.g. closed/broken pipe) aborts the delete with this wrapper.

Source

Thrown at cmd/kops/delete.go:90

	cmd.AddCommand(NewCmdDeleteInstance(f, out))
	cmd.AddCommand(NewCmdDeleteInstanceGroup(f, out))
	cmd.AddCommand(NewCmdDeleteSecret(f, out))
	cmd.AddCommand(NewCmdDeleteSSHPublicKey(f, out))

	return cmd
}

func RunDelete(ctx context.Context, factory *util.Factory, out io.Writer, d *DeleteOptions) error {
	// We could have more than one cluster in a manifest so we are using a set
	deletedClusters := sets.NewString()

	for _, f := range d.Filenames {
		var contents []byte
		var err error
		if f == "-" {
			contents, err = ConsumeStdin()
			if err != nil {
				return fmt.Errorf("reading from stdin: %v", err)
			}
		} else {
			contents, err = factory.VFSContext().ReadFile(f)
			if err != nil {
				return fmt.Errorf("reading file %q: %v", f, err)
			}
		}

		sections := text.SplitContentToSections(contents)
		for _, section := range sections {
			o, gvk, err := kopscodecs.Decode(section, nil)
			if err != nil {
				return fmt.Errorf("parsing file %q: %v", f, err)
			}

			switch v := o.(type) {
			case *kopsapi.Cluster:
				options := &DeleteClusterOptions{

View on GitHub (pinned to 4c8573c808)

Solutions

1. Ensure something actually pipes data in: `cat resources.yaml | kops delete -f -`.
2. Check the upstream command in the pipeline succeeded before kOps consumed stdin.
3. If you meant to read a file, pass the filename instead of `-`.

Example fix

// before (no stdin in CI)
kops delete -f -
// after
kops delete -f resources.yaml

// ensure stdin has data when using '-'
const stdinData = fs.readFileSync(0, 'utf8'); // throws immediately if stdin empty/closed
if (!stdinData.trim()) throw new Error('No manifest provided on stdin');

try {
  runKops(['delete', '-f', '-']);
} catch (e) {
  if (/reading from stdin/.test(e.message)) {
    console.error('stdin read failed; pipe a manifest in or use -f <file>.');
  }
  throw e;
}

When it happens

Trigger: Running `kops delete -f -` where stdin is closed, is not piped at all, or the producer process terminates mid-stream (broken pipe / EOF condition surfaced as an error).

Common situations: Shell pipelines like `kops get cluster -o yaml | kops delete -f -` where the upstream command fails; interactive shells with no stdin attached; CI runners without stdin allocated when `-` is accidentally used instead of a filename.

Related errors

• reading file %q: %v
• must specify %q label with cluster name to create instanceGr
• error querying cluster %q: %v
• cluster %q not found
• instanceGroup %q already exists

AI-assisted analysis of kubernetes/kops@4c8573c808 (2026-09-05). Data as JSON: /api/errors/9b6b75b93fbd0627. Report an issue: GitHub.