Sustained latency above two seconds on Mirror 02 from 18:14 to 19:01 UTC. Root cause traced to upstream rendezvous churn on the descriptor publishing path. Rebalanced to Mirror 01 and 03 transparently. No login or escrow disruption.
Daily timestamp signing job failed silently on the custodian side. Detected after community report. Re-signed and republished. Mirrors remained operational throughout, only the freshness indicator was stale.
Tor exit consensus failure caused descriptor unreachability for Mirror 01. Mirror 02 and 03 remained healthy and absorbed full traffic. Resolved by descriptor republish after consensus stabilised.
Three lookalike clearnet domains burned through community reports. Updated phishing watch on the Chronicle. No platform impact, communications only.
Subkey rotation for support, vendor, and dispute desks. Master key fingerprint 0A9D unchanged. Ceremony witnesses signed cross-signatures, public key bundle republished on each mirror.
Three impact levels. Minor means a single mirror affected, no buyer or vendor session disruption. Major means two or more mirrors affected simultaneously, or a single mirror down for more than 30 minutes. Maintenance covers planned key ceremonies, signed announcements, and security communications that change nothing user-facing but do require operator action.
Resolution times are reported as wall-clock from first detection to verified recovery. We do not distinguish between detection lag and actual outage start because the user experience is identical, the market is unreachable either way. The reported duration is the true outage window.
Every incident entry is signed under the master fingerprint 0A9D. The archive itself is also signed monthly by the audit observer key, witnessed by independent custodians, and republished on each mirror under /announcements/audit. If you care about the chain of custody, the cross-signatures are the proof.
Across the last 90 days, the dominant incident class is minor latency spikes on individual mirrors during EU peak hours. The cause is upstream Tor exit congestion and is not addressable from our side. The mitigation is having three mirrors. None of these incidents have caused user-facing disruption because traffic absorbs across the roster.
The second most common class is signature freshness lag, usually triggered by a custodian schedule conflict. The mitigation is the cross-signed observer key, which catches stale signatures within minutes and pages the next custodian on rotation.
Genuine outages, defined as descriptor unreachability for an entire mirror for more than 15 minutes, average less than one per quarter and have never affected more than one mirror at once.
No mailing list. No webhook. No push channel.
The archive is published here and republished on each mirror under /announcements. If you want to follow incidents, bookmark this page. If you want signed alerts for high-impact events, the security mailing inside the market core sends them to vendors and active buyers.
Anything claiming to be a Nexus Market status feed on Telegram, Discord, RSS, or anywhere else is impersonation. Report it to the security desk.