Skip to content

Troubleshooting

github-actions[bot] edited this page Oct 10, 2026 · 2 revisions

Troubleshooting

The page loads but no logs show up

  1. Look at the log view first. Problems reading a source are shown there as [frontail] <source>: <reason>:
    • cannot open ... Permission denied: the user running frontail can't read the file. ls -l it; add the user to the file's group (often adm), or fix the file's mode. In Docker the process is a non-root user, so a root-only host file fails the same way.
    • cannot open ... No such file or directory: wrong path. frontail keeps waiting and starts streaming if the file appears later.
    • ssh exited with code N; reconnecting in Ns: the SSH connection failed or dropped (see below).
  2. Check the server's own output. Run it in the foreground and read stderr.
  3. Behind a reverse proxy? See "Log view empty behind a proxy".

"Syslog isn't working" / /var/log/syslog doesn't exist

Many current distributions (Fedora, Arch, recent Debian and Ubuntu) log to the systemd journal and have no /var/log/syslog; /var/log/messages may exist but be readable only by root. Use the journal:

frontail --journal

The user needs access to the journal (member of systemd-journal or adm). To follow only some services: --journal-unit sshd --journal-unit nginx.

Log view empty behind a proxy

Two usual causes:

  • The proxy doesn't forward WebSocket upgrades. Add proxy_http_version 1.1; and the Upgrade/Connection headers (Reverse proxy).
  • The proxy rewrites Host, so frontail's cross-site check rejects the browser. Keep the original host (proxy_set_header Host $host;) or add --allowed-origin https://your-public-host.

SSH sources

  • Permission denied (publickey...): ssh never prompts (BatchMode=yes). Set up key auth for the user running frontail, and make sure the host key is already trusted (ssh host true once by hand).
  • Could not resolve hostname, Connection timed out: network or DNS. The connect timeout is 10 s.
  • ssh exited with code N; reconnecting in Ns: the session dropped and frontail is retrying with backoff (up to once a minute). If it never recovers, check the same causes as above (keys, host key, network).
  • Ports, keys and jump hosts come from ~/.ssh/config.

Colors are off

  • Make sure you're on 2.31 or later: 2.28 to 2.30 had a bug that turned colorizing and indentation off by default.
  • The Colors button in the sidebar overrides the server default per browser; check it's active.
  • Lines that already contain ANSI codes keep their own colors and skip format detection.

Container logs

  • Failed to spawn docker: the engine isn't installed or isn't on PATH (use --container-engine podman for Podman).
  • In the Docker image, mount the socket: -v /var/run/docker.sock:/var/run/docker.sock:ro.

Other

  • Arguments needed, use --help: give at least one source (a file, -, -C, --journal, --ssh) or a files list in the config file.
  • Port already in use: change -p.
  • Node too old: frontail needs Node 22.12+ (the Docker image and binaries bundle their own).
  • Unknown option "x" in config file: config keys must match option names; see Configuration file.
  • Still stuck? Open an issue with your version (frontail -V), the command line (without secrets) and what the log view says.

Clone this wiki locally