What this tool does
Measures real round-trip latency to a host by timing TCP connection setup from our server — the accurate, browser-accessible equivalent of ICMP ping, since neither browsers nor most hosted web servers can send raw ICMP echo packets.
Why TCP instead of ICMP?
Sending a traditional ICMP ping requires raw socket access, which is intentionally restricted on shared and cloud hosting for security reasons. Timing a TCP handshake instead gives you the same real network round-trip measurement — the three-way handshake still has to physically travel to the target and back — without needing elevated privileges.
Reading the results
Min/avg/max show the spread of round-trip times across four attempts; a wide spread often points to congestion or an unstable path rather than a consistently slow one. Packet loss here means a connection attempt timed out completely with no response at all, distinct from a fast "connection refused," which still confirms the host is reachable.
Troubleshooting high latency or packet loss
A ping time under about 20ms is excellent (typical for a well-routed connection to a nearby server), 20-100ms is normal for most regional and cross-country connections, and anything consistently above 150-200ms will start to feel laggy for real-time use like gaming or video calls. What matters more than the raw number is consistency -- a stable 80ms is usually fine, while a result that swings between 20ms and 300ms points to congestion somewhere on the path, often at a shared Wi-Fi access point, an oversaturated ISP link during peak hours, or an overloaded router hop.
Packet loss almost never originates at the destination server -- it is far more commonly a local Wi-Fi interference issue, a failing cable or port, or your ISP dropping packets under load. If you see loss on this tool but pages still load normally in your browser, that is expected: browsers retry failed TCP connections automatically, while a ping-style test reports every individual attempt.