STATUS LOG / READ ONLY

Is Torzon Down? Live Mirror Status

This log tracks which Torzon nodes answered the last probe and when. Onion hosts drop and come back often, so a row can read Checking one minute and live the next. The full addresses live on the home directory; this page is only the pulse.

4/5Answered last probe
~1hSample lag
24hProbe window
1Known-dead node
PROBE LOGlast run 2026-08-19

Working links and last probe

node.01torzongu…mjtdzqdChecking3m ago
node.022vycvnjt…m3qe5ydChecking3m ago
node.033t6qm72c…adhmuqdChecking4m ago
node.0445bgvcy7…cved4adChecking4m ago
node.055anlnmvt…pt3nhadDead1d ago

Fragments only. The complete addresses sit in the signed directory on the home page. A row here is a probe state, never a safety rating, and it can trail reality by an hour.

WHY A NODE LAGSpropagation

Why a Torzon node can read down and answer minutes later

An onion service is not offline the way a normal site is. It keeps re-publishing a descriptor to the Tor network on a timer, and your client has to fetch that descriptor before it can connect. When the upload slips under load, the two sides fall out of step. That gap is what a Checking row is really showing.

WHY A NODE LAGSSVConion serviceHSDIRpublishes descriptorRELAYyour Tor pathYOUfetch + connect
01 · publishThe service uploads a fresh descriptor to the responsible directory relays on a schedule.
02 · lookupYour client fetches that descriptor before it can build a circuit to the onion.
03 · slipUnder load or attack the upload lags, so one path sees it and another does not.
04 · readA Checking row means the last lookup timed out. It is not proof of a seizure.
READING THE LOGplain read

What down really means here

Not working usually means slow or overloaded, not seized. A single stalled node says little. Work down the log to the next one that answered. The signal worth reacting to is different: when every row reads Dead for a full day, treat the whole set as gone and wait for a fresh signed list.

Is Torzon down, or is it just my connection?

Check two things before you decide. If one node reads Checking while another answers, the market is up and that single node is lagging. If your own circuit has gone stale, close the tab, build a new Tor identity, and retry. A real outage looks different: every row dead, held for hours, not minutes.

What counts as a working link here?

A working link is a full onion from the signed directory that answered a recent probe and still matches the fingerprint. A live pill is half the test. An address that loads but fails the signature is a working clone, which is worse than a dead node, because it looks like success.

Honest limit: this log is a snapshot, not a heartbeat. We sample on a loose schedule, so do not read a live row as proof that an address is genuine. That is still the fingerprint's job.

STATUS NOTEShow to read it

Reading the Torzon status grid without fooling yourself

Why does a node read Checking and not Down?

Because the probe cannot tell a slow onion from a gone one in a single try. Checking means the last lookup timed out and the next is pending. It is an honest maybe, not a verdict. We never paint a row green on a hunch, so a status you see is the raw probe result and nothing kinder.

How often does this log refresh?

On a loose schedule measured in minutes to an hour, not per second. That lag is on purpose. Hammering an onion to draw a live dot would only pile load onto a service already fighting for descriptor space. Read a row as a recent sample, then confirm the address yourself.

Torzon is not working for me. What is the order of moves?

Rebuild your circuit first. Then try the next node on the grid that answered. Then pull a fresh address from the signed directory in case the one you saved has gone stale. If all of that fails and every node stays dead for a day, stop and wait for a new signed list rather than hunting for a link elsewhere.

Does a dead node mean the market was seized?

Rarely. A single dead node is routine churn. Even the whole grid going quiet points to load or a maintenance window far more often than a takedown. Treat a seizure claim as a rumour until a signed message says otherwise, because panic is exactly what a phishing page wants from you.

FULL DIRECTORY

Get the complete addresses

Copy full onions from the signed directory, then match each one against the fingerprint before you connect.

Open directory