Skip to content

TCP port monitoring

TCP port monitoring allows you to check whether a TCP service is accepting connections - a database, an SMTP or SSH server, a message broker, a game server, or any other host:port endpoint - by opening a TCP connection to it.

How does it work?

Kuvasz monitors your services by periodically opening a TCP connection to the configured host and port. For each check it measures:

  • Reachability (whether the connection could be established within the timeout)
  • Connect latency (the time it took to establish the connection)

If the connection cannot be established within the configured timeout, or - when configured - the connection takes longer than the optional latency threshold, the monitor is marked as DOWN and you will be notified through your configured notification channels.

These two settings cover different failure modes and are meant to complement each other. The timeout is a hard ceiling that decides whether the port is reachable at all, while the latency threshold is a lower bar (set well below the timeout) that flags connections which succeed but are too slow. For example, with a 5000 ms timeout and a 500 ms latency threshold, a connect that takes 800 ms is still reachable but breaches your latency SLA, so the monitor is marked DOWN. Leave the latency threshold unset if you only care about reachability.

What can be configured?

  • interval for uptime checks
  • hostname or IP address to connect to
  • TCP port to connect to (1-65535)
  • connection timeout (1-30000 milliseconds)
  • optional connect-latency threshold that triggers a DOWN event
  • consecutive failure count threshold

Configuration

Please refer to the Managing TCP monitors section of the documentation for more information on how to configure TCP port monitoring.