Warm-standby: promote runbook + single mobile-triggerable recovery path #262

New issue

Closed

opened 2026-06-07 05:27:13 +00:00 by coilysiren · 1 comment

coilysiren commented

2026-06-07 05:27:13 +00:00

Owner

Warm-standby follow-up from #99. Write the promote runbook: restore the latest home-1 snapshot onto ser8, bring up a single-node k3s server there, and repoint workers. RTO = restore time + manual trigger.

Plus a single mobile-triggerable recovery path — one action Kai can fire from her phone (Android Claude client → kai-server/ser8, or a one-shot script) that executes the promote end to end. Must be runnable when home-1 is fully dark.

Restore is sqlite/kine-shaped, not etcd (see snapshot-ship issue). Depends on snapshot-ship existing and on #258.

Warm-standby follow-up from #99. Write the promote runbook: restore the latest home-1 snapshot onto ser8, bring up a single-node `k3s server` there, and repoint workers. RTO = restore time + manual trigger. Plus a **single mobile-triggerable recovery path** — one action Kai can fire from her phone (Android Claude client → kai-server/ser8, or a one-shot script) that executes the promote end to end. Must be runnable when home-1 is fully dark. Restore is sqlite/kine-shaped, not etcd (see snapshot-ship issue). Depends on snapshot-ship existing and on #258.

coilysiren added the

label

2026-06-17 07:05:48 +00:00

coilysiren commented

2026-06-17 08:22:34 +00:00

Author

Owner

Backlog burndown 2026-06-17: closing low-priority (P3/P4) to bring the open count to a manageable level. Nothing lost — reopen if this resurfaces. Batch tag: burndown-2026-06.

Backlog burndown 2026-06-17: closing low-priority (P3/P4) to bring the open count to a manageable level. Nothing lost — reopen if this resurfaces. Batch tag: `burndown-2026-06`.

coilysiren

2026-06-17 08:22:34 +00:00