Public: anyone can read this; the crew can change it.
Maintenance from the forge: request a shutdown, keep a page up, and come back by itself #9453
Landed in pull #9639 · 2 runs · run
2 earlier comments
Parked by the crew: the check stayed red through 2 automatic tries (pull #9557 closed). Planner: split this smaller, or rewrite what done looks like; a person can also say what to try.
Reviewed pull #9639 at its head: ticked An admin presses "Prepare for maintenance" (and the API does the same): releases hold, the crew pauses, a fresh backup is taken, the forge drains, and the host can power off. The page says plainly what is about to stop.; Part of going forge-only (#7667): this works with no Kubernetes underneath.. left open: While it's off, visitors see a calm maintenance page served from somewhere else (the droplet or Cloudflare), not an error. — scripts/standby.ts serves the page, but it is started by hand and nothing repoints the tunnel or pushes it to Cloudflare.; On power-on, everything returns by itself and verifies: the forge, apps, tunnel, runner and the nightly timers. The admin is told when it's back and what's healthy. — The return verifies the database, hold, dispatcher, apps on record and nightly lines, but it does not probe the tunnel or the runner, and nothing is pushed to the admin.; A standby can take over reads, or the whole forge, from a second machine with enough room, so a planned move isn't an outage. State the size needed. — The size is stated, but nothing replicates data to a second machine, so there is no takeover..
Left open after pull #9639 landed: 3 unticked items remain (While it's off, visitors see a calm maintenance page served from somewhere else (the droplet or Cloudflare), not an error.; On power-on, everything returns by itself and verifies: the forge, apps, tunnel, runner and the nightly timers. The admin is told when it's back and what's healthy.; A standby can take over reads, or the whole forge, from a second machine with enough room, so a planned move isn't an outage. State the size needed.). The planner files each as its own issue and closes this one; a person may close it, tick what is done, or say what is left. Until then the crew does not pick it again.
The owner, 2026-11-07, moving the server rack: "get everything onto the droplet and power off the vxrail... i expect everything to move back over seamlessly. i want these controls in the forge. i should be able to just request a shutdown or something. this is also our opportunity to get off of the platform and go solely forge."
What happened by hand: releases held, the forge scaled to zero, the vxrail powered off, then on boot the forge scaled back and was verified. The forge was unreachable for the whole move, because its data (38 GB, apps 13 GB) and its tunnel live only on the vxrail, and the droplet has 16 GB of disk and 3 GB of free memory.
Done when: