anomalyco/sst · error

failed to copy lambda bridge: %v

Error message

failed to copy lambda bridge: %v

What it means

During PythonRuntime.Run, after syncing artifacts, the runtime copies the lambdaric_python_bridge.py from the platform dist (dist/python-runtime/index.py) into the build output when it's missing or older than the source. Failure of copyFile is wrapped as "failed to copy lambda bridge: %v". Without the bridge file, the worker would lack its entrypoint, so this is fatal.

Source

Thrown at pkg/runtime/python/python.go:151

func (r *PythonRuntime) Run(ctx context.Context, input *runtime.RunInput) (runtime.Worker, error) {
	isLegacyLayout, err := r.syncArtifactsIfNeeded(input)
	if err != nil {
		slog.Error("failed to sync artifacts",
			"functionID", input.FunctionID,
			"error", err)
		return nil, fmt.Errorf("failed to sync artifacts: %v", err)
	}

	// Copy lambda bridge to artifact directory if missing or outdated
	lambdaBridgePath := filepath.Join(input.Build.Out, "lambdaric_python_bridge.py")
	sourceBridgePath := filepath.Join(path.ResolvePlatformDir(input.CfgPath), "/dist/python-runtime/index.py")

	dstInfo, dstErr := os.Stat(lambdaBridgePath)
	srcInfo, srcErr := os.Stat(sourceBridgePath)
	if dstErr != nil || (srcErr == nil && srcInfo.ModTime().After(dstInfo.ModTime())) {
		if err := copyFile(sourceBridgePath, lambdaBridgePath); err != nil {
			return nil, fmt.Errorf("failed to copy lambda bridge: %v", err)
		}
	}

	projectRoot := path.ResolveRootDir(input.CfgPath)

	var handlerPath string
	var workingDir string

	if isLegacyLayout {
		// Use relative handler since workingDir is the artifact directory
		handlerPath = r.adjustHandlerForFlattenedLayout(input.Build.Handler)
		workingDir = input.Build.Out
	} else {
		// Modern layout: run from source with PYTHONPATH
		handlerPath = input.Build.Handler
		workingDir = projectRoot
	}

View on GitHub (pinned to a0bd20f762)

Solutions

1. Build the platform first (`bun run build:platform` or `bun run setup`) so dist/python-runtime/index.py exists.
2. Check the wrapped %v error: if 'no such file', the platform dist is missing; if 'permission denied', fix output dir permissions.
3. Verify the build output directory (input.Build.Out) is writable by the current user.
4. Clear the stale artifact dir and rebuild to force a fresh bridge copy.

Example fix

// before: fresh clone, platform not built
// failed to copy lambda bridge: ... no such file

// after
$ bun run setup && sst dev

// shell, before dev/deploy
test -f <platform-dir>/dist/python-runtime/index.py && echo OK || echo 'platform not built: run bun run build:platform'

try { await deploy() } catch (e) {
  if (String(e).includes("failed to copy lambda bridge")) {
    await buildPlatform(); // regenerate dist/python-runtime/index.py
    await deploy();
  }
  throw e;
}

When it happens

Trigger: os.Stat found the destination bridge missing/stale and copyFile failed: source bridge absent (platform not built — platform/dist/python-runtime/index.py missing), destination directory not writable, or disk full.

Common situations: Fresh clone without running `bun run setup`/platform build, so ResolvePlatformDir(...)/dist/python-runtime/index.py doesn't exist; deploying in a container with a read-only platform dir; permissions broken on .sst output after running commands as root vs user.

Related errors

• failed to sync artifacts: %v
• default Python Dockerfile not found at %s: %w
• failed to find package archive: %w
• no package archive found for %s (tried patterns: %s-*.whl, %
• failed to process archive %s: %w

AI-assisted analysis of anomalyco/sst@a0bd20f762 (2026-08-30). Data as JSON: /api/errors/05208c53524760c2. Report an issue: GitHub.