First steps

Use this checklist for a new Thruk installation. Choose the installation method before you change the configuration.

Choose an installation

Method	Use it when
Packages	You want Thruk on an existing monitoring server or a separate web server.
OMD	You want a monitoring stack with Thruk and other monitoring tools.
Source	You need custom paths or dependencies and can maintain the installation yourself.

Use the packages for your distribution. Do not install a package built for another distribution or release.

Production web server

The development server on port 3000 is not a production server. Use a configured web server and FastCGI or PSGI deployment.

Configure access

1. Open the Thruk URL supplied by your installation. Standalone packages normally use http://your-host/thruk/.

2. Configure HTTPS before you send credentials over the network.

3. For standalone packages, set a password for the thrukadmin account:

   htpasswd /etc/thruk/htpasswd thrukadmin

4. Sign in and check the displayed username.

OMD uses site-specific paths and accounts. Follow its site setup instructions instead of applying standalone paths unchanged.

See web server integration and access control for authentication details.

Connect one backend

First, confirm that the monitoring core provides Livestatus. Thruk displays core data. Installing Thruk does not create monitoring checks.

For a standalone package, put local settings in /etc/thruk/thruk_local.conf. Replace the socket path below with the socket supplied by your monitoring core:

<Component Thruk::Backend>
  <peer>
    name = local
    type = livestatus
    <options>
      peer = /var/run/live
    </options>
  </peer>
</Component>

1. Give the Thruk process access to the socket.
2. Reload the service that runs Thruk after configuration changes, as required by your deployment.
3. Open the host or service list.
4. Check that the backend is available and its objects appear.

Do not expose an unencrypted Livestatus connection to an untrusted network. For remote sites, see TLS and HTTP backends.

Check a normal user's access

Test with a non-administrator account before you give users access.

1. Confirm that the username matches a monitoring contact or the intended role mapping.
2. Check which hosts and services the user can see.
3. Check whether the user can submit commands.
4. For read-only users, confirm that command submission is unavailable.

Authentication alone does not assign monitoring permissions. See access control and CGI authorization settings.

Add features after the basic checks pass

• Create dashboards for operators.
• Configure SLA reports if you need scheduled availability reports.
• Use the REST API for automation.
• Plan backups and upgrades before production use.

If objects do not appear, check logs and the FAQ.