Ping Test

Ping uses the Internet Control Message Protocol, defined for IPv4 in RFC 792 and for IPv6 in RFC 4443. The sender transmits an Echo Request (ICMP type 8, or type 128 in ICMPv6) carrying an identifier, a sequence number and a payload; the target copies them into an Echo Reply (type 0, or 129) and sends it back. The time between sending a request and receiving the matching reply is the round-trip time (RTT). RFC 1122 requires every host to implement an echo responder, but nothing stops administrators from filtering it, which is why a silent ping is not proof that a server is down.

RTT is dominated by physics and path. Light travels through optical fibre at roughly 200,000 km per second, so every 100 km of cable adds about 1 ms to a round trip before any queuing or processing. A server on the same continent commonly answers in 10 to 60 ms, while transatlantic paths sit around 70 to 100 ms and Europe to Australia often exceeds 250 ms. Rising RTT under load usually means congested queues somewhere on the path, sometimes called bufferbloat, and variation between replies (jitter, shown here as standard deviation) matters as much as the average for voice and video calls.

Packet loss is reported as the share of requests without a reply. Routers and firewalls often rate-limit ICMP or treat it as low-priority traffic, so a lost echo does not always mean that TCP traffic would be lost. Conversely, a target behind a cloud security group that blocks ICMP returns 100 percent loss while its web server works perfectly. The TTL value in each reply is a useful side clue: operating systems start from typical defaults such as 64 (Linux, macOS), 128 (Windows) or 255 (many network devices), and the difference to the received value roughly indicates how many routers the reply crossed.

Anycast and CDNs add another nuance. Many large services announce the same IP from dozens of locations, so a ping reaches the nearest edge node rather than the origin server, and the measured latency describes that edge. When you need to test the origin, ping its direct address or use an HTTP check against the origin hostname.

This tool first resolves the hostname and refuses private, loopback and other reserved targets to prevent server-side request forgery, preferring the IPv4 address when both families exist. It then sends four echo requests from our server, waiting up to two seconds for each reply, and returns the counts of sent and received packets, the loss percentage and min, average, maximum and standard deviation of the RTT. Zero loss is a pass, any loss is a warning, and 100 percent loss is reported as unreachable. Keep in mind that the latency is measured from our infrastructure, not from your own connection.