abiosoft/colima · error

process signal(0) returned error: %w

Error message

process signal(0) returned error: %w

What it means

The daemon liveness probe sends signal 0 (existence check, no actual signal delivered) to the pid read from the pid file. An error means errno ESRCH (no process with that pid — stale pid file, daemon died) or EPERM (process exists but belongs to another user — pid was recycled after a reboot or by another daemon). The wrapped syscall error distinguishes the two.

Source

Thrown at cmd/daemon/daemon.go:134

	}

	// check if process is actually running
	p, err := os.ReadFile(info.PidFile)
	if err != nil {
		return fmt.Errorf("error reading pid file: %w", err)
	}
	pid, _ := strconv.Atoi(string(p))
	if pid == 0 {
		return fmt.Errorf("invalid pid: %v", string(p))
	}

	process, err := os.FindProcess(pid)
	if err != nil {
		return fmt.Errorf("process not found: %v", err)
	}

	if err := process.Signal(syscall.Signal(0)); err != nil {
		return fmt.Errorf("process signal(0) returned error: %w", err)
	}

	return nil
}

const (
	pidFileName = "daemon.pid"
	logFileName = "daemon.log"
)

func Info() struct {
	PidFile string
	LogFile string
} {
	dir := dir()
	return struct {
		PidFile string
		LogFile string

View on GitHub (pinned to c3a5f9184d)

Solutions

1. Check the wrapped errno: ESRCH means stale pid file — remove dir()/daemon.pid and run `colima daemon start`; EPERM means another user's process holds the pid — likewise delete the stale pid file
2. Inspect the daemon log (dir()/daemon.log) for why the daemon exited
3. Avoid mixing sudo and non-sudo invocations of colima daemon

Example fix

# before
$ colima daemon status
process signal(0) returned error: no such process

# after
$ rm .../daemon.pid   # remove stale pid file
$ colima daemon start
$ colima daemon status

// probe liveness yourself before relying on daemon status
func daemonAlive(pidFile string) bool {
    b, err := os.ReadFile(pidFile)
    if err != nil {
        return false
    }
    pid, err := strconv.Atoi(strings.TrimSpace(string(b)))
    if err != nil || pid <= 0 {
        return false
    }
    proc, _ := os.FindProcess(pid)
    return proc.Signal(syscall.Signal(0)) == nil
}

// distinguish stale-pid (ESRCH) from foreign-owner (EPERM)
if err := status(); err != nil {
    var errno syscall.Errno
    if errors.As(err, &errno) {
        switch errno {
        case syscall.ESRCH: // stale pid file -> safe to remove
        case syscall.EPERM: // pid recycled to another user -> also stale for us
        }
    }
}

When it happens

Trigger: Daemon exited or crashed but left daemon.pid behind; host rebooted and the stale pid now maps to a live process of another user (EPERM); daemon started under sudo so a non-root status check hits EPERM; pid recycling on long-lived hosts.

Common situations: Skipping `colima daemon stop` in favor of kill/reboot; daemon started with sudo and queried as regular user (or vice versa); stale daemon.pid surviving a colima upgrade or profile switch; an unrelated process adopting the same pid after reboot.

Related errors

• pid file not found: %w
• invalid pid: %v
• cannot make dir: %w
• error sending sigterm to daemon: %w
• process not found: %v

AI-assisted analysis of abiosoft/colima@c3a5f9184d (2026-08-15). Data as JSON: /api/errors/5cf30f693140bfa0. Report an issue: GitHub.