Port monitoring that catches the outage your website hides.
The website still loads while the service behind it is down — and you hear about it from users first. Uptimia checks the exact port from 171+ locations, as often as every 30 seconds, and pages you once several regions confirm it's really down. Databases, mail, SSH, game servers, custom daemons: if it listens on a port, it's watched.
Monitoring Logs
every 30 s · rotating locations
Tokyo14:32:18—Down
London14:32:16—Down
New York14:32:14—Down
Frankfurt14:32:11—Down
Amsterdam14:31:4111 msUpNew York14:31:1194 msUpFrankfurt14:30:416 msUpResponse Time
avg 178 ms · last 7 daysThe morning Redis stopped answering
Nothing in the application logs, nothing on the status page. The cache port simply refused the connection at 09:47:11, and every request that needed it started queueing behind a timeout.
Every port your stack listens on
Databases, mail, SSH, message brokers, game servers, custom daemons — any TCP or UDP port, plus ping and DNS, checked from outside your network.
Confirmation, alerting and evidence
Paged only when several regions can't connect
One checkpoint failing to connect never pages you. Uptimia re-tests the port from other regions — only once they agree does an incident open. Recovery stays fail-safe: one good connection clears it.
- Confirmation you control — 1 to 3 independent regions must fail to connect, default 3
- One outage, one page — a single incident opens, not an alert from every checkpoint that saw it
- Your alert delay — hold a confirmed failure for up to 30 minutes before it pages anyone
agree
The right on-call knows within seconds
A confirmed failure alerts the channels your team already lives in. If nobody reacts, the ladder pulls in the next responder, and a recovery notice closes it out.
- Escalation ladders that keep paging the next on-call until someone acknowledges — no login needed
- Maintenance windows and pause — a planned restart never wakes anyone
- Recovery notice with the outage duration when the port answers again
Which phase of the connection broke
Each incident keeps exactly what the failing checkpoints saw — DNS, connect, TLS, send or receive — so you start the fix from facts, not by trying to reproduce it at 2 a.m.
- The exact phase that failed — connect refused, TLS error, or a socket that opened but never answered
- Expected banner vs what came back — when you set a send/expect string
- Traceroute to the port — see the hop where the packets stopped, then share it as PDF, HTML or a public link
send PING\r\n expect +PONG got — (no data)
9 ae-3.sin 12 ms
14 be2.ams 214 ms
15 * * * lost
Checked from 171+ locations in 70+ countries
Checkpoints on six continents show when a port answers from Toronto but is refused from Amsterdam — a firewall rule or route that only bites some of your users. Use the whole network or pick the regions your clients connect from.
- As fast as every 30 seconds — a flapping port can't hide between checks
- Connection time per checkpoint — charts break down by region
- Every probe IP is published — whitelist them through your firewall in one pass
Name a host and a port.Hear it from us, not from your error logs.
Every protocol, every region, every alert channel — free for 30 days, and none of it is a paid add-on.
How port monitoring works
Live in under a minute — nothing to install; every check runs from our probe network.
Name the host and port
Pick TCP or UDP; frequency and locations come pre-filled — change anything, or nothing.
Decide who gets paged
Choose the channels and the people — escalation ladders and quiet hours are optional.
Get alerted in seconds
A confirmed failure opens one incident and alerts every chosen channel.
Also included
Full REST API
Stand up, edit, pause and clone port checks straight from CI or scripts, with per-user API keys.
TLS-wrapped mail checks
Tick encryption on the mail ports that require it — IMAPS, SMTPS, POP3S. TCP checks take the same flag through the API.
Scheduled reports
Availability and connection-time reports, carrying your own logo and colours.
Public status pages
Tell customers a service is healthy — and keep them posted when it isn't.
Maintenance windows
A scheduled restart never pages the on-call.
Every monitor type in one account
Port checks sit beside uptime, SSL, speed, DNS and heartbeat monitors — same contacts, groups and roles.
Alerts where your team already works
One failure, delivered everywhere — the same contacts and channels whether it's a port, a page or a certificate.
Microsoft Teams12 channels, one contact list — configure it once and every check type reuses it.
Browse the full integrations directory →What is port monitoring?
Port monitoring is an automated service that repeatedly opens a connection to a specific TCP or UDP port and checks that the service behind it answers correctly — usually every 30 seconds to a few minutes, from many locations at once. When the connection is refused or the reply is wrong, it alerts you by email, SMS or chat, so a dead service is caught before it takes your application down.
How does port monitoring work?
Each check opens a real socket, optionally sends a probe string and matches the reply, building your uptime and connection-time history.
Confirm first, then alert
3 of 3 regions agree → the incident opens and the alerts go out
Which port was that, again?
Most outages a team actually cares about live on a handful of well-known ports. Uptimia watches any of them — TCP or UDP, plain or wrapped in TLS — on the same cadence as your website.
Free tool: see which ports answer on a host →| Port | Service | Typical check |
|---|---|---|
| 22 | SSH | TCP connect + banner |
| 5432 | PostgreSQL | TCP connect |
| 3306 | MySQL | TCP connect |
| 6379 | Redis | TCP · send PING, expect PONG |
| 587 | SMTP submission | TCP + STARTTLS |
| 53 | DNS | UDP or TCP resolve |
Port monitoring FAQ
01What is port monitoring?+
02How does a port check work?+
03What's the difference between port monitoring and ping?+
04Which ports and protocols can I monitor?+
05Can it check the service actually works, not just that the port is open?+
06How do you prevent false alarms?+
07Can I monitor a service that's only reachable internally?+
08What do I get when a port goes down?+
09How do I get alerted when a port goes down?+
10Is there a free plan?+
11Does every check run from all 171+ locations at once?+
12Which plans include port monitoring?+
Start monitoring your ports today.
Name a host and port, pick your channels — be the first to know when a service stops answering.