What causes "daemon exited during startup." in multica-ai/multica?

Open the Full log path printed in the error — it contains the complete startup attempt.

daemon exited during startup.

Error message

daemon exited during startup.

What it means

Constructed by daemonStartupFailureError in cmd_daemon.go when the spawned daemon child process exited before ever reporting ready, and the log-tail inspection did not match either recognized failure signature (auth token rejected, or network-unreachable). This is the fallback branch: the final error embeds an exit-code variant when available, a noise-filtered excerpt (WRN/ERR lines plus raw crash output like Go panics), and the full log paths so the user can diagnose directly.

Source

Thrown at server/cmd/multica/cmd_daemon.go:796

	}
	excerpt := filterDaemonLogNoise(lines, 5)
	// Crash output is already raw (panics, the final cobra error line) — no
	// noise filter, just cap it like the structured excerpt.
	if len(crashLines) > 5 {
		crashLines = crashLines[len(crashLines)-5:]
	}
	excerpt = append(excerpt, crashLines...)
	if len(excerpt) > 0 {
		b.WriteString("\nLog output:")
		for _, line := range excerpt {
			fmt.Fprintf(&b, "\n  %s", line)
		}
	}
	fmt.Fprintf(&b, "\nFull log: %s", logs.logPath)
	if len(crashLines) > 0 {
		fmt.Fprintf(&b, "\nCrash output: %s", logs.errLogPath)
	}
	return errors.New(b.String())
}

// filterDaemonLogNoise drops DBG/INF log lines from an excerpt, keeping
// WRN/ERR entries and plain (non-slog) lines such as the final error the
// child printed before exiting, capped to the last max lines.
func filterDaemonLogNoise(lines []string, max int) []string {
	out := make([]string, 0, len(lines))
	for _, line := range lines {
		if strings.Contains(line, " DBG ") || strings.Contains(line, " INF ") {
			continue
		}
		out = append(out, line)
	}
	if len(out) > max {
		out = out[len(out)-max:]
	}
	return out
}

View on GitHub (pinned to 2c0912b6ec)

Solutions

Open the Full log path printed in the error — it contains the complete startup attempt.
Check the Crash output path (daemon.err.log) for the Go panic or cobra 'Error:' line; that names the root cause.
Re-run `multica login` and confirm the server URL is correct even if the heuristic missed the auth failure shape.
Move aside/reset the daemon state directory if logs show corruption, then retry `multica daemon start`.
If the panic persists, capture the crash tail and file it with the daemon version string.

Defensive patterns

Strategy: fallback

Try / catch

err := waitForDaemonReady(ctx, child, logs)
if err != nil {
  return daemonStartupFailureError(logs, err, profile, serverURL) // already includes log paths + filtered excerpt
}

Prevention

Read the printed Full log and Crash output paths first — they contain the root cause for this fallback branch.
Run `multica login` after token rotation to avoid the auth-rejection variant.
Keep daemon state and log directories writable by the daemon user.
Smoke-test `multica daemon start` after every upgrade in a scratch profile.

When it happens

Trigger: `multica daemon start` spawns the child which crashes before the ready signal: panic in daemon initialization, invalid config file, port already bound, missing permissions on state dirs, or any startup failure not matching the auth/network heuristics.

Common situations: Corrupted daemon state directory; config with an invalid value after an upgrade; another process holding the daemon port; binary/arch mismatch after a partial update; permissions changed on the log/state dirs.

Related errors

AI-assisted analysis of multica-ai/multica@2c0912b6ec (2026-08-15). Data as JSON: /api/errors/c39d3b391a3288e8. Report an issue: GitHub.