TCP Port Check (Ping)
Check if a TCP port on a public host is open and measure connection latency.
Loading TCP Port Check (Ping)… If nothing happens, please enable JavaScript.
Frequently asked questions
What data does this tool send to your servers?
Why does this tool use TCP instead of ICMP like a real ping?
The check times out even though the server is up — what does that mean?
What do the different result statuses mean?
Are there rate limits on how many checks I can run?
How does this compare to running 'ping' or 'nc' in a terminal?
Can I check private or internal IP addresses?
Does this tool log the hosts I check for analytics or advertising?
What port numbers are commonly associated with standard services?
Many firewalls block ICMP — does that mean ping is useless?
About TCP Port Check (Ping)
Ping is one of the most fundamental network diagnostic tools, used to determine whether a remote host is reachable and to measure the round-trip time of a network connection. The classic 'ping' command uses ICMP (Internet Control Message Protocol) Echo Request and Echo Reply messages, which travel at the IP layer below TCP and UDP. Network engineers rely on it daily to confirm that routers, servers, and appliances are online, to measure latency baselines, and to detect packet loss indicating a congested or unreliable link. The round-trip time (RTT) reported by ping is one of the key metrics in every network performance investigation.
Sysadmins use ping to quickly verify that a server came back online after a reboot, or to confirm that a firewall rule change didn't inadvertently block connectivity. Developers test API server reachability and latency. Web hosting customers check whether their server IP is responsive after a reported outage. Security teams use port-level reachability checks to verify that firewall rules are enforcing the correct allow and deny policies — confirming that port 22 (SSH) is closed to the internet while port 443 (HTTPS) is open, for example. Even gamers use latency measurements to choose the lowest-latency game server region.
Because standard ICMP ping requires elevated operating system privileges not available in shared hosting environments, this tool performs a TCP three-way handshake connect to the port you specify rather than sending ICMP packets. The result tells you the same essential information — is the host accepting connections? — while also being more operationally useful for service-specific checks: you can test port 80 for HTTP, port 443 for HTTPS, port 25 for SMTP, port 22 for SSH, and so on. The hostname is resolved server-side using public DNS, and private IP ranges (RFC 1918), loopback, and cloud metadata endpoints are blocked to prevent abuse. Latency is measured as the time from the TCP SYN to the completion of the handshake.
When interpreting results, a successful connection with low latency (under 50 ms for nearby servers) indicates the port is open and the service is responding. A 'connection refused' error means the host is reachable but nothing is listening on that port, which is different from a timeout — a timeout usually means a firewall is silently dropping packets rather than actively rejecting them. If you are troubleshooting a service outage, try both the expected port and port 443/80 to distinguish between a service crash and a network-level block. Consistently high latency (above 200 ms) is worth investigating with traceroute for routing anomalies.
From 'Submarine Sonar' to Network Debugging in 40 Years
The ping command was written by Mike Muuss in December 1983 as a debugging tool to investigate strange behaviour on a network he was managing. He named it after the sound a submarine sonar makes when it bounces a sound pulse off an object — the analogy being that a network ping sends a signal and listens for an echo to determine if something is 'out there.' Muuss wrote the original implementation in a single evening, and the program was so useful that it spread rapidly across the ARPANET and eventually became a standard utility on every operating system. Muuss died in a car accident in 2000, never having trademarked or commercialised the tool that bears his sonic metaphor.
ICMP, the protocol that traditional ping uses, was defined in RFC 792 by Jon Postel in September 1981. It operates at the IP layer rather than the transport layer (TCP/UDP), making it the internet's built-in signalling and error-reporting mechanism. ICMP is used for more than ping — it also carries traceroute's TTL-exceeded messages, path MTU discovery signals, and the 'destination unreachable' responses that tell applications their connections have failed. The TTL (Time to Live) field in IP packets was specifically designed to prevent routing loops from causing packets to circulate forever; each router decrements it by one, and ICMP reports when it hits zero.
The transition from ICMP ping to TCP-based reachability checks reflects a broader shift in network philosophy over the decades. Early internet design assumed trusted, cooperative participants and made network infrastructure maximally transparent. Modern networks treat external probes as potential attack vectors, leading to firewalls that block ICMP by default and security groups that only allow traffic on specific application ports. This is why operations teams increasingly measure application-layer reachability — can I complete an HTTPS handshake? — rather than raw network reachability, since a server that responds to application requests is functionally up even if it silently drops all ICMP probes.