argoproj/argo-workflows · warning

failed waiting for debug-pause-after marker: %w

Error message

failed waiting for debug-pause-after marker: %w

What it means

When ARGO_DEBUG_PAUSE_AFTER=true, argoexec pauses after the container command exits and waits for the marker file /var/run/argo/ctr/<container>/after to be created, letting a developer inspect state. This error is thrown when that wait is interrupted — almost always because the surrounding context was cancelled or its deadline expired before the marker appeared.

Source

Thrown at cmd/argoexec/commands/emissary.go:279

					if err != nil {
						logger.WithField("containerName", containerName).WithError(err).Error(innerCtx, "failed to terminate/kill container")
					}
				}()

				break
			}
		}

		return osspecific.Wait(command.Process)
	})
	logger.WithError(cmdErr).Info(ctx, "sub-process exited")

	if os.Getenv("ARGO_DEBUG_PAUSE_AFTER") == "true" {
		// User can create the file: /ctr/NAME_OF_THE_CONTAINER/after
		// in order to break out of the wait and release the container from
		// the debugging state.
		if waitErr := file.WaitForCreate(ctx, varRunArgo+"/ctr/"+containerName+"/after"); waitErr != nil {
			return fmt.Errorf("failed waiting for debug-pause-after marker: %w", waitErr)
		}
	}

	exitCode = exitCodeFromErr(cmdErr, exitCode)

	if containerName == common.MainContainerName {
		for _, x := range template.Outputs.Parameters {
			if x.ValueFrom != nil && x.ValueFrom.Path != "" {
				if err := saveParameter(ctx, template, x.ValueFrom.Path); err != nil {
					return err
				}
			}
		}
		for _, x := range template.Outputs.Artifacts {
			if x.Path != "" {
				if err := saveArtifact(ctx, template, x.Path); err != nil {
					return err
				}

View on GitHub (pinned to 35bff19146)

Solutions

1. Exec into the pod and create the marker: kubectl exec <pod> -c <container> -- touch /argo/ctr/<containerName>/after.
2. Extend or remove activeDeadlineSeconds / termination grace while debugging.
3. Unset ARGO_DEBUG_PAUSE_AFTER when not actively debugging.
4. Do not delete/terminate the workflow while paused.

Example fix

// before (pod deleted while paused)
ARGO_DEBUG_PAUSE_AFTER: "true"  # no one creates the marker
// after
# in another terminal, before the deadline:
kubectl exec -it <pod> -c main -- touch /argo/ctr/main/after

// Before relying on the pause, confirm the env and that you can exec in:
// kubectl exec <pod> -c <ctr> -- sh -c 'test -w /argo/ctr && echo ok'

try {
  await waitPodTermination(pod)
} catch (e) {
  if (String(e).includes('debug-pause-after')) {
    // create the marker or release the workflow:
    // kubectl exec <pod> -c <ctr> -- touch /argo/ctr/<ctr>/after
  }
}

When it happens

Trigger: ARGO_DEBUG_PAUSE_AFTER env set to "true" but nobody creates the /ctr/<name>/after file before the context is cancelled (workflow terminated, pod deleted, debug timeout elapsed).

Common situations: Developer enables the debug pause but forgets to exec into the pod and touch the marker file; the pod's grace period or the workflow's activeDeadlineSeconds kills it first; kubectl exec session fails so the file is never created.

Related errors

• failed to read container args file %s: %w
• failed to unmarshal container args: %w
• failed to read template: %w
• failed to start command: %w
• failed to create emissary: %w

AI-assisted analysis of argoproj/argo-workflows@35bff19146 (2026-09-03). Data as JSON: /api/errors/da41ac14dba9bbf8. Report an issue: GitHub.