Skip to content
RU

Open ports in 2026: 1,042 checks run by users themselves

TL;DR. We did not scan third-party hosts on our own initiative — instead we counted 1,042 checks that users ran themselves, choosing the host and the port set.The result is calmer than the common image: in 51.1% of cases not one port is open from the set examined, and in a further…

We did not scan third-party hosts on our own initiative — instead we counted 1,042 checks that users ran themselves, choosing the host and the port set.

The result is calmer than the common image: in 51.1% of cases not one port is open from the set examined, and in a further 28.6% one or two. Five or more ports is the top 3%. Below: the full distribution and an honest account of how this sample is biased.

Check your host & ports →

What exactly we counted, and what we did not do

We did not scan third-party hosts on our own initiative. That matters: mass port-checking of sites that never asked for it is not research but interference.

Instead we took 1,042 checks that users ran themselves through our tool, choosing the host and the port set. So this is not a slice of the internet but a slice of what people check when they have a reason to: almost always their own server, or one that raised a suspicion.

Standard sets were used: most often seventeen common ports, sometimes eight or seven. The summary records how many of them answered.

Half the time nothing is open

Open portsChecksShare
053251.1%
217917.2%
111911.4%
3979.3%
4646.1%
5131.2%
8121.2%
6101.0%
770.7%
1660.6%
9 and 1230.3%

More than half the checks found no open port at all in the set examined. A further 28.6% found one or two — usually the web itself: 80 and 443.

Five or more ports open occurs in 3% of checks. Sixteen of seventeen, in six.

The conclusion these numbers support: the typical host somebody chooses to check is closed. The image of a server with ports flung wide describes not the norm but a rare tail of the distribution.

Why these numbers say nothing about the internet

The bias here is strong and runs both ways at once, so it is worth stating in full.

Toward closed: people most often check their own server, one they look after. Such a host is by definition configured better than average.

Toward open: the second common reason to run a check is a suspicion that something is exposed. Those checks land in the right-hand tail.

The port set is fixed. Seventeen common ports are not all 65,535. A service on a non-standard port is invisible to our numbers.

A closed port does not mean a secure one. The check sees whether a port answers from our address. A firewall admitting only known addresses will read as "closed" — correctly. And an open 443 with a vulnerable application behind it reads as "open" and says nothing about the risk.

What these numbers do give is a benchmark: if more than four ports of a standard set are open on your host, you are in the top 3% and it is worth checking whether all of them are needed.

Where to start if more ports are open than expected

  1. Find out what is listening. On the server itself ss -tlnp shows the process behind each port. Half of unexpected findings are services brought up by a package manager and never turned off.
  2. Separate "listening on 0.0.0.0" from "listening on 127.0.0.1". A database bound to the local address is invisible from outside and will not appear in a check. One bound to all addresses will — and that is the commonest genuine finding.
  3. Close with a firewall, not by stopping the service. The service may be needed, while a firewall rule survives a restart and a package upgrade.
  4. Check both protocols. Our set looks at TCP; a UDP service does not enter it.

To check your own host, use the port scanner; to see what a web service returns, the header check.

ICMP PingHost availability and latency
Port ScannerOpen TCP port detection
LatencyResponse time in milliseconds
Packet LossPercentage of dropped packets

Why teams trust us

ICMP+TCP
check protocols
14
key ports scanned
2,014
checks in 30 days
3
regions

How it works

1

Enter IP or domain

2

ICMP packets sent

3

Latency & packet loss shown

How Do Ping and Port Scanning Work?

Ping sends ICMP packets to a host and measures response time. Port scanning checks which TCP ports are open and accepting connections — helping diagnose serviceavailability issues.

Configurable Ping

Choose packet count (3, 4, 6, 10). Stats: min/avg/max latency and packet loss.

Common Port Scanner

Check 14 key ports: HTTP, HTTPS, SSH, FTP, SMTP, MySQL, PostgreSQL, and more.

Cloud-Based Check

Testing from our server — see site availability from outside, not just your local network.

Uptime Monitoring

Need constant monitoring? Create a monitor — checks every minute with notifications.

Who uses this

DevOps

availability diagnosis

Network engineers

TCP port scanning

Developers

connection debugging

SRE

basic health check

Common Mistakes

ICMP blocked = server is downMany servers block ICMP. Ping fails but site works — check ports instead.
High ping = server problemLatency depends on geography. 150ms between continents is normal, not an error.
Closed ports — cause for alarmClosed ports of unused services are good. Unnecessary open ports are a risk.
One check = sufficientNetworks are unstable. A single timeout ≠ a problem. Check multiple times or set up monitoring.

Best Practices

Combine ping and port checksPing shows host availability, ports show specific service availability. Use both.
Check from different locationsThe problem may be local. A cloud test shows the real picture.
Close unused portsEvery open port is a potential attack vector. Keep only necessary ports open.
Set up monitoringManual checks do not scale. Set up automated monitoring with notifications.

Get more with a free account

Ping check history, host availability monitoring and downtime alerts.

Sign up free

Learn more

Frequently Asked Questions

What if my server is in the "no auth" list?

1) Immediately bind service to 127.0.0.1 (or VPC internal IP), 2) firewall drop port, 3) add auth, 4) audit access log for predators. In exactly that order.

Why does Beget hit "22% risky"?

Shared hosting: one IP serves 100+ clients. If one client has a misconfigured Redis — the whole IP registers as "risky" in our metrics.

How do I quickly check my ports?

Enterno Ping + Port Checker — enter your domain, see which ports are reachable from the internet. Or: nmap -sT yourdomain.com.

Which ports are safe to open at all?

Only what your app needs: 443 (web), 22 (SSH, mandatory key-auth, no password), optionally 80 (redirect). DB/cache/queue — always bind 127.0.0.1 or private network.

Try the live tool that powered this guide

Free plan — 10 monitors, checks every 5 min, no card required. Upgrade for 1-minute interval and multi-region monitoring.