Status // Rolling 30 days
Mirror Performance and Incident Log
This section records how each mirror has behaved over the past month. We track availability, median round-trip time, and a rolling thirty-day uptime figure for every address. Data refreshes on a fixed schedule rather than on demand, so there is a small lag between an event and its appearance here.
Current readings
| No | Address | Status | Last check | |
|---|---|---|---|---|
| 01 | blackops27m32abqvbhnyswgazqawxqbznbzkkkv5sjo7gve2ndpsdad.onion | Online | Nov 10, 2026 | |
| 02 | blackops5l63qnwnmlnsfvtlu66md3x3vp3fdtpna42eq7ozujr67hid.onion | Online | Nov 10, 2026 | |
| 03 | blackops6kignp3eddmvqcfkjzf6qr6haxbmkypc2xtqlnhuu4ak4dqd.onion | Online | Nov 10, 2026 |
We publish these numbers so you can pick the most responsive entry point at any given hour. If all three show similar metrics, any of them will serve you equally well. See which mirror to use for the decision rules.
How we measure
Status is determined by a scripted probe that attempts a TCP connection and expects an HTTP response within a defined timeout. Median latency is the middle value from a series of timed requests spread across the measurement window. Thirty-day uptime is the percentage of scheduled probes that succeeded out of the total number attempted.
These are operational readings, not service-level agreements. A 99.7 percent figure means roughly five failed probes per month, not a guaranteed five-minute outage cap. Treat the numbers as a trend indicator, not a contract.
Recent incidents
Three events required attention during the current reporting period. None affected user data, and none demanded action beyond patience or a simple retry.
Connection resets on mirror 01
The first mirror experienced fourteen minutes of intermittent connection resets. Pages either failed to load or dropped mid-request. The issue resolved itself once the upstream node recovered, and no user intervention was necessary beyond refreshing the page.
Latency spike on mirror 03
The third mirror ran two hours above its normal latency baseline. Content still loaded, but each request took several times longer than usual. No data was lost; sessions remained valid throughout the spike.
Scheduled rotation on mirror 02
The second mirror underwent a planned address rotation. The old .onion string was retired and the replacement became live within the same maintenance window. Users holding bookmarks to the old address simply switched to the new one; no credentials changed hands.