nexusb2l7fmqnefwphyy7m5zjhlkytlbo7qbb5lu5dlczr3azgii2gyd.onionnexusma2iqgauqqvjcgds4ckv5xbf272tkfagq4epojjhsgleqpwxiqd.onionnexusabcd6tyfhdwilyitaqiri6tisj2v2hueyjuj6qkvd6azvi5tuqd.onionRun the check before you sign in. Open these in Tor Browser only. When the login screen loads, compare the onion printed on the page against your address bar, reading from the end. If they differ, close the tab.
This one is further along than 0xF0 and that matters. Your client found the record, contacted a relay the service nominated, and the service itself did not pick up.
Why this is a more hopeful error
Reaching this point proves the descriptor is live and recent. Something published it, and the introduction points listed in it are real relays that accepted the connection. The only step that failed is the last one, where the service should have answered. Services that have genuinely gone away stop publishing and produce 0xF0 instead.
The usual causes
- The service is restarting. A restart takes seconds to minutes and the descriptor outlives it, which produces exactly this error in the gap.
- The service is saturated and dropping new introductions to protect the ones it already has. Under a flood this is deliberate behaviour rather than a fault.
- One introduction point in the descriptor has gone stale. The client will pick a different one on the next attempt, which is why retrying sometimes works instantly.
What to do
Retry once straight away, since a different introduction point may be chosen. If that fails, wait two or three minutes and try again rather than hammering. Repeated attempts during a flood are indistinguishable from the flood itself and get you nothing.
If a second mirror answers immediately, use it. Getting on with the order matters more than working out which address is healthiest.
What it does not mean
It is not a sign the market has been seized or has run off. Both of those look like a permanent 0xF0 across every address, not a disconnect on one. It also does not mean your account or session is affected, because nothing reached the application layer.
How to read the result
Nothing was wrong. A stale introduction point got replaced. No action needed and no pattern to read into it.
The address is genuinely struggling. Switch to another mirror and check back later rather than sitting on it.
Treat it as a network wide event affecting the service. Wait it out. Do not go looking for alternative addresses.
Other reachability signals
Onionsite Not Found
What the 0xF0 error means when a Nexus mirror will not open, why retrying rarely helps, and which of the three addresses to try next.
ReachabilityUnable to connect to the onionsite
The 0xF3 error sits between a lookup failure and a real page, and it usually says more about your circuit than about the Nexus mirror.
ReachabilityNo error page, just a spinner that gives up
When a Nexus mirror produces no error at all and the tab simply spins, the useful information is in how long it spins before quitting.
Procedures that apply here
How to take an honest reading of a mirror
A repeatable way to measure whether a Nexus mirror is slow, broken or fine, without tooling and without fooling yourself.
MethodChoosing which of the three mirrors to use
The three Nexus addresses reach the same market. How to pick between them sensibly and why the fastest one today may not be tomorrow.
IncidentWhat to do when none of the three answer
How to tell a service wide outage from a problem on your side, and the one action that turns an inconvenience into a loss.