TCP port monitoring
Check that any service port is open and reachable.
A TCP monitor opens a connection to a host and port and verifies the service accepts it. Built for anything that doesn't speak HTTP.
How it works
UptimeAlien attempts a TCP handshake to host:port at your interval and measures the connect time; failures and timeouts mark the monitor down.
When to use it
- Databases reachable only from the inside: PostgreSQL 5432, MySQL 3306, Redis 6379
- SMTP, IMAP and POP3 ports, where a failure sends no alert mail because the mail server is exactly what is gone
- SSH on machines you rarely touch, where an outage would otherwise surface at your next login
- Game and voice servers, where users complain to each other rather than to you
- Custom TCP services with no HTTP interface, which would otherwise have no check at all
What you get
Connection success, connect latency, uptime ratios and incidents.
Example
db-primary.acme.com:5432
Frequently asked questions
What is TCP port monitoring?
A TCP monitor opens a connection to a host and port and checks that the handshake succeeds. It confirms a service is listening and reachable, without needing to speak the protocol running on that port.
Which services can I monitor this way?
Anything that listens on a TCP port: databases like PostgreSQL or MySQL, mail servers, Redis, message brokers, SSH, game servers or an internal service behind a load balancer.
What is measured besides reachability?
The connect time is recorded as the response time, so you can see a port getting slower to accept connections before it stops accepting them altogether.
What happens if the port is closed?
The connection attempt fails or times out and the check counts as failed. After the configured number of consecutive failures the monitor goes down and alerts fire.