Root cause confirmed: the outage was caused by our hosting provider, Contabo. At 12:12 UTC they powered off one of our servers from their side, without warning, as part of their emergency security patching of all their VPS hosts (critical Linux kernel vulnerability CVE-2026-99999). Our server logs show a power-off signal sent by their infrastructure; no action was taken on our side and nothing failed in our platform itself.

Contabo's notice: https://contabo-status.com/ ("Important security update: patching of all VPS/VDS hosts", 15 Oct 2026). Contabo says more host restarts may happen while their patching continues. Our setup now survives the loss of that server, so a further restart on their side should not cause downtime.

Resolved. Between 12:12 and 12:20 UTC, the app and the API were unavailable: logins, API calls and uploads failed.

Cause: one of our servers went down unexpectedly. It was hosting our internal DNS service and the primary database, so the rest of the platform could no longer reach the database.

Fix: we moved the DNS service to a healthy server and switched the database to its replica. No data was lost. Scheduled posts that failed during this window can be retried from the app.

To prevent this from happening again, the DNS service now runs on two separate servers.