What causes "rc-service restart failed: {}" in zeroclaw-labs/zeroclaw?

Check service state: `rc-service zeroclaw status`; if stopped, start it manually with `sudo rc-service zeroclaw start` and read the error output

zeroclaw-labs/zeroclaw · critical · anyhow::Error

rc-service restart failed: {}

Error message

rc-service restart failed: {}

What it means

Thrown by maybe_restart_managed_daemon_service on Linux when an OpenRC service is detected (/etc/init.d/zeroclaw exists and `rc-service zeroclaw status` reports running) but `rc-service zeroclaw restart` exits non-zero. The prior status check succeeded, so a failed restart likely leaves the service stopped or in a degraded state; the surrounding caller surfaces a manual restart command.

Source

Thrown at crates/zeroclaw-channels/src/orchestrator/mod.rs:8791

        return Ok(true);
    }

    if cfg!(target_os = "linux") {
        // OpenRC (system-wide) takes precedence over systemd (user-level)
        let openrc_init_script = PathBuf::from("/etc/init.d/zeroclaw");
        if openrc_init_script.exists()
            && let Ok(status_output) = Command::new("rc-service").args(OPENRC_STATUS_ARGS).output()
        {
            // rc-service exits 0 if running, non-zero otherwise
            if status_output.status.success() {
                let restart_output = Command::new("rc-service")
                    .args(OPENRC_RESTART_ARGS)
                    .output()
                    .context("Failed to restart OpenRC daemon service")?;
                if !restart_output.status.success() {
                    let stderr = String::from_utf8_lossy(&restart_output.stderr);
                    anyhow::bail!("rc-service restart failed: {}", stderr.trim());
                }
                return Ok(true);
            }
        }

        // Systemd (user-level)
        let home = directories::UserDirs::new()
            .map(|u| u.home_dir().to_path_buf())
            .context("Could not find home directory")?;
        let unit_path: PathBuf = home
            .join(".config")
            .join("systemd")
            .join("user")
            .join("zeroclaw.service");
        if !unit_path.exists() {
            return Ok(false);
        }

View on GitHub (pinned to 88bb9c8533)

Solutions

Check service state: `rc-service zeroclaw status`; if stopped, start it manually with `sudo rc-service zeroclaw start` and read the error output
Inspect /etc/init.d/zeroclaw and confirm its command/program path matches the installed binary; reinstall the init script if stale
Run the zeroclaw binary in the foreground with the same config to find why startup fails, and fix the config error
If privileges were the issue, run config-reload operations as a user permitted to control the service, or switch to the user-level systemd path

Example fix

# before: restart fails after binary moved
# rc-service restart failed: /usr/bin/zeroclaw: No such file or directory

# after
sudo rc-update del zeroclaw default
sudo cp contrib/openrc/zeroclaw /etc/init.d/   # or edit program path
sudo rc-update add zeroclaw default
sudo rc-service zeroclaw start

Defensive patterns

Strategy: fallback

Validate before calling

if !Path::new("/etc/init.d/zeroclaw").exists() {
    return Ok(()); // no OpenRC service installed
}
let status = Command::new("rc-service").args(["zeroclaw", "status"]).output()?;
if !status.status.success() { return Ok(()); } // not running; skip restart path

Try / catch

if let Err(e) = maybe_restart_managed_daemon_service() {
    eprintln!("restart failed ({e:#}); recover with `sudo rc-service zeroclaw start`");
    // alert: prior stop means the daemon may now be DOWN
}

Prevention

Run config-mutating CLI commands as a user permitted to control the system service
Reinstall the init script whenever the binary path changes
Validate config in the foreground before any operation that restarts the service

When it happens

Trigger: A config-mutating operation (e.g. allowlist save) triggers auto-reload on an Alpine/Gentoo-style OpenRC host where /etc/init.d/zeroclaw exists and the service is up, and `rc-service zeroclaw restart` fails — typically the init script's stop/start commands erroring (bad binary path, config crash on start, or insufficient root privileges for the invoking user).

Common situations: Running the CLI as a non-root user on a system-wide OpenRC service; init script references a deleted/moved zeroclaw binary; daemon config became invalid so the start half of restart fails; OpenRC supervise-daemon timeout during slow shutdown.

Related errors

AI-assisted analysis of zeroclaw-labs/zeroclaw@88bb9c8533 (2026-08-23). Data as JSON: /api/errors/90ae8a3c66d86941. Report an issue: GitHub.