What causes "failed to run OS detection command: %s" in hashicorp/packer?

Check communicator logs immediately before this error for the underlying SSH/WinRM cause.

failed to run OS detection command: %s

Error message

failed to run OS detection command: %s

What it means

detectRemoteOS error when comm.Start(ctx, cmd) fails to launch the OS detection command ('uname -s -m' over SSH, 'echo %PROCESSOR_ARCHITECTURE%' over WinRM) on the remote machine. This indicates the communicator could not start the command at all (connection/protocol-level failure), as distinct from the command running and exiting non-zero.

Source

Thrown at provisioner/hcp-sbom/provisioner.go:384

	}

	// Run detection command based on communicator
	var cmd *packersdk.RemoteCmd
	if connType == "winrm" {
		cmd = &packersdk.RemoteCmd{
			Command: "echo %PROCESSOR_ARCHITECTURE%",
		}
	} else {
		cmd = &packersdk.RemoteCmd{
			Command: "uname -s -m",
		}
	}

	var stdout bytes.Buffer
	cmd.Stdout = &stdout

	if err := comm.Start(ctx, cmd); err != nil {
		return "", "", fmt.Errorf("failed to run OS detection command: %s", err)
	}

	cmd.Wait()

	if cmd.ExitStatus() != 0 {
		return "", "", fmt.Errorf("OS detection command exited with status %d", cmd.ExitStatus())
	}

	output := strings.TrimSpace(stdout.String())
	log.Printf("OS detection output: %s", output)

	// Parse output
	var osType, osArch string
	if connType == "winrm" {
		osType = "Windows"
		osArch = strings.ToLower(output) // AMD64, ARM64, etc.
	} else {
		parts := strings.Fields(output)

View on GitHub (pinned to eb36e3c3e4)

Solutions

Check communicator logs immediately before this error for the underlying SSH/WinRM cause.
Verify SSH/WinRM connectivity (ssh into the guest or test WinRM) with the same credentials the builder uses.
Increase communicator timeouts / add retry settings if the connection is flaky.
Note the build itself continues: Provision logs 'Failed to detect remote OS' and skips SBOM generation, so this is non-fatal to the image build.

Defensive patterns

Strategy: try-catch

Validate before calling

// pre-check communicator reachability before the provisioner runs
client, err := ssh.Dial("tcp", fmt.Sprintf("%s:%d", host, port), sshConfig)
if err != nil {
	return fmt.Errorf("guest not reachable over SSH before OS detection: %w", err)
}
session, err := client.NewSession()
if err != nil {
	return fmt.Errorf("cannot open SSH session (command start would fail): %w", err)
}
session.Close()

Try / catch

if err := provisioner.Provision(ctx, ui, comm, data); err != nil {
	if strings.Contains(err.Error(), "failed to run OS detection command") {
		// non-fatal for the build: log, check SSH/WinRM config, optionally retry
		ui.Error("OS detection failed; skipping SBOM generation")
		return nil
	}
	return err
}

Prevention

Validate SSH/WinRM credentials, port, and host key settings in the builder config
Increase communicator timeouts on slow-booting guests
Verify the guest shell exists and is permitted for the login user
Remember Provision treats this as non-fatal (build continues without SBOM) — fix connectivity to restore SBOM generation

When it happens

Trigger: comm.Start returns an error in detectRemoteOS, reached from Provision when generatedData lacks usable OSType/OSArch — e.g. SSH session cannot allocate/start the remote command, WinRM command dispatch fails, or the connection dropped.

Common situations: SSH connection dropped or not fully established when the provisioner runs; WinRM endpoint misconfigured (port/auth/HTTPS); guest shell unavailable or restricted; firewall/timeout killing the session mid-build.

Related errors

AI-assisted analysis of hashicorp/packer@eb36e3c3e4 (2026-09-05). Data as JSON: /api/errors/c95474d4fb27e0ce. Report an issue: GitHub.