Live Ping Monitor
Intermittent faults need a time series rather than a single reading. Sampling runs only while the tab is open and visible, and uses HTTP requests rather than ICMP.
Live Ping Monitor
Loading the test…
How this test works
Probe: one HTTP request per second to /cdn-cgi/trace of the nearest node, timed with Resource Timing when available. Timeout: 5 s per probe. Retention: the chart shows the last 5 minutes; a run stops after 30 minutes. The run finishes when the tab becomes hidden.
How to read the result
Statistics shown: current, minimum, average and maximum latency, jitter, lost percent, spike count and probe count. Lost probes are counted apart and never become 0 ms. A spike is a probe above the selected limit: 50, 100, 200, 300 or 500 ms, default 200. The chart scale is about 1.25 times the visible maximum.
What is typical
The loss percent here counts failed HTTP probes, which is not the same as packet loss on the network. The spike limit presets are fixed values. The chart spans at least 30 s and grows up to the 5 minute window.