GoogleContainerTools/skaffold · error

retrieving image config: %w

Error message

retrieving image config: %w

What it means

RetrieveConfigFile resolves an image's config by first trying a local Docker daemon; if no local daemon is available it falls back to RetrieveRemoteConfig (go-containerregistry remote). If both paths fail, the error is wrapped as "retrieving image config: %w". Called by RetrieveWorkingDir and RetrieveLabels, it means neither the local daemon nor the remote registry could supply the image metadata.

Source

Thrown at pkg/skaffold/docker/image_util.go:46

func RetrieveConfigFile(ctx context.Context, tagged string, cfg Config) (*v1.ConfigFile, error) {
	if strings.ToLower(tagged) == "scratch" {
		return nil, nil
	}

	var cf *v1.ConfigFile
	var err error

	localDocker, err := NewAPIClient(ctx, cfg)
	if err == nil {
		cf, err = localDocker.ConfigFile(context.Background(), tagged)
	}
	if err != nil {
		// No local Docker is available
		cf, err = RetrieveRemoteConfig(tagged, cfg, v1.Platform{})
	}
	if err != nil {
		return nil, fmt.Errorf("retrieving image config: %w", err)
	}

	return cf, err
}

func RetrieveWorkingDir(ctx context.Context, tagged string, cfg Config) (string, error) {
	cf, err := RetrieveConfigFile(ctx, tagged, cfg)
	switch {
	case err != nil:
		return "", err
	case cf == nil:
		return "/", nil
	case cf.Config.WorkingDir == "":
		log.Entry(context.TODO()).Debugf("Using default workdir '/' for %s", tagged)
		return "/", nil
	default:
		return cf.Config.WorkingDir, nil
	}

View on GitHub (pinned to a1189de023)

Solutions

1. Ensure the image exists: either build/pull it locally (`docker pull <tag>`) or push it to an accessible registry.
2. Authenticate for the remote registry so RetrieveRemoteConfig can read the manifest (`docker login`, `gcloud auth configure-docker`).
3. Fix the image reference in the config — verify with `docker manifest inspect <tag>`.
4. Start the local Docker daemon so the local-daemon path succeeds instead of falling back to remote.

Example fix

// before: base image only exists locally, daemon down
$ skaffold dev
// after
$ sudo systemctl start docker && docker pull gcr.io/my-proj/base:latest
$ skaffold dev

// ensure the reference resolves either locally or remotely
if !daemon.ImageExists(ctx, tagged) {
    if _, err := remote.Head(tagged); err != nil {
        return fmt.Errorf("image %q neither local nor pullable: %w", tagged, err)
    }
}

cf, err := RetrieveConfigFile(ctx, tagged, cfg)
if err != nil && strings.Contains(err.Error(), "retrieving image config") {
    return nil, fmt.Errorf("image %q not found locally or in registry; push it or fix the tag: %w", tagged, err)
}

When it happens

Trigger: Calling RetrieveConfigFile(ctx, tagged, cfg) when the local daemon lacks/errs on the image AND RetrieveRemoteConfig fails — image not present locally, not published to any registry, remote registry auth missing, or malformed image reference.

Common situations: Referencing a base image built locally but never pushed while the daemon is unreachable, private base image with no `docker login` / credential helper configured, typo in the image name/tag in skaffold.yaml, offline environment with no local image.

Related errors

• getting auth config: %w
• getting auth config for %q: %w
• %s %q: %w
• pulling image from repository: %w
• getting image: %w

AI-assisted analysis of GoogleContainerTools/skaffold@a1189de023 (2026-09-05). Data as JSON: /api/errors/e9ee9d0718384498. Report an issue: GitHub.