Brief restart of all services (network layer maintenance) Sunday 15th November 2026 16:58:01


The container network layer is being restarted (network proxy maintenance). Every service is unavailable for a few minutes and comes back on its own.

What happened: at 16:55 the network proxy that carries all public traffic crashed under an unusually heavy automated crawl of the Mastodon instance (about 100 requests per second from a rotating pool of addresses). Every service was unreachable from 16:55 to 17:00; the proxy was restarted and all services were verified healthy at 17:00. No data was affected. Measures against the crawl are being taken.

All services are back after 2 min 7 s. Started by hand after the restart: nginx-proxy, nginx-proxy-gen.

Related Mastodon account