Per-mirror uptime, last 30 days
Per-mirror uptime for the last 30 days with sparklines, incident count, MTTR (mean time to recovery) and MTBF (mean time between failures). Higher uptime and lower MTTR at the top.
How uptime is measured
A distributed prober fleet checks every mirror every 5 minutes. Success is a 200 response inside 60 seconds. Failure is a timeout, a 5xx, or a connection reset. Failures get re-checked from a different Tor exit before being counted, to filter noise from a single bad circuit.
What MTTR and MTBF mean here
MTTR is the average length of an outage window. When a mirror flips to "down", the timer starts. When it flips back to "up", the timer stops. MTTR is the mean across 30 days.
MTBF is the average interval between outages. If a mirror has 4 incidents in 30 days (720 hours), MTBF is roughly 180 hours. Higher MTBF means fewer surprises.
Interpreting the sparkline
Each bar covers roughly one day. Green means high uptime, cyan is normal, red is a bad day. A single red bar is a Tor network hiccup, not a mirror problem. Several red bars in a row usually means a real outage.