A TCP check attempts to establish a connection to a configured public hostname and port. It is useful for services that do not expose an HTTP page, but a successful connection is only a reachability signal.
Set up a TCP target
When adding a monitored target, choose TCP Port, enter a target name, public host and port from 1 to 65535. WebCheck stores the target as a tcp:// host-and-port address and applies the configured timeout and schedule.
What the check records
The TCP transport resolves the hostname, rejects restricted addresses, attempts the connection and records connection time when successful. Failures are classified as DNS resolution failure, connection refused, timeout or another connection failure.
Examples
- A game server listening on a public port can be checked for connection reachability.
- A mail or API service can be checked at its public service port.
- A closed port produces connection refused even when the host itself is reachable.
- A firewall drop may appear as a connection timeout rather than a refused connection.
Important limitation
An open TCP port does not prove that the application protocol is healthy, authentication works or a user transaction succeeds. Use an HTTP health endpoint when application-level health is what you need to measure.