What is UDP port monitoring?
Game servers, NTP time servers, VoIP, SIP, syslog and many IoT protocols talk over UDP instead of TCP. A UDP port monitor sends one datagram to the host and port on a schedule and waits for an answer. If nothing comes back within 10 seconds, the port is closed, or the answer isn't what you expected, the monitor goes down and your team is alerted.
It works on any port from 1 to 65535, with no agent to install. Each reply's round-trip time is recorded on the response-time chart.
Why UDP needs a request and a reply
TCP has a handshake, so a port check can tell an open port from a closed one just by connecting. UDP has none: a datagram sent to a running service and one sent to a dead server both disappear silently. The only proof a UDP service works is an answer to something it understands.
So you tell the monitor what to send, such as a game server status query or an NTP request, and optionally what the reply must contain. Use plain text for text protocols, or tick hex for binary ones. If the server sends back ICMP port unreachable, the check fails at once instead of waiting for the timeout.
Fewer false alarms
UDP has no retransmission, so a single lost packet can look like an outage. When a check fails, Spot Downtime retries it after 2 seconds and checks from other regions at the same time; the monitor only goes down when most of them agree. On Starter and Pro you can run every check from Amsterdam or Singapore, or from several regions at once, to watch a server from where its players are.
Checks come from the public internet and private addresses are blocked, so the service must be reachable from outside. For DNS servers, a DNS monitor is the better fit: it checks the records that come back, not just that the server replied.

