Classic ping uses ICMP, which many servers and clouds block. A TCP ping opens a real connection to the web port and times the handshake, so it shows whether the service is reachable and how far away it is from our server.
Table of Contents
Interpreting the numbers
Consistent times under a few milliseconds mean the host is in or near the same data center as our server; 70–150 ms is typical across the Atlantic. Big variation between attempts points at congestion or an overloaded host, and failed attempts with “connection refused” mean the port is closed or firewalled.
For a full path analysis run mtr or traceroute from your own machine; this test measures reachability and latency only.
TCP ping at a glance
Official documentation: Windows Server documentation, AlmaLinux wiki, Linux man pages.
Related guides: A basic RMM monitoring policy for small-business endpoints: disk, patching and antivirus · Account quotas show “unlimited”: fixquotas and XFS/ext4 quota repair on cPanel · Check domain controller health with dcdiag and repadmin (what each test means).
Frequently asked questions
Why use TCP instead of ICMP?
Firewalls and cloud providers often drop ICMP, so an ICMP ping can fail while the website works. A TCP handshake tests the service you actually care about.
Can I test other ports?
This public tool is limited to 80 and 443. Use nc -zv host port from a server you control for other ports.
Does a successful TCP ping mean the website works?
It means the port accepts connections. Use the website down checker to see the HTTP response.