Recover: VPS1 / de-james-vps1
No signal importantSecond VPS, in a different provider region from VIN (Hetzner nbg1 against VIN's fsn1), running its own complete Coolify instance. Registered here because it is the estate's only second compute failure domain that has been proven to execute work. It is shared: another operator's watchdog agent has root on it, and it already terminates 13 public hostnames that are not this estate's. Nothing this estate depends on runs here yet.
- Owner
- homelab-operator
- Target RTO
- 1h
- Last heartbeat
- never
- Dependent entities
- 0
Recovery readiness20/100
Procedure
Read fully before acting.- Nothing this estate depends on runs here, so this node being down is not an estate incident. Do not treat it as one.
- What its loss does cost is the second failure domain: while it is down, every HA claim that names it is a plan again. Say so in the incident rather than working around it.
- Administrative access is Tailscale SSH as root, governed by the tailnet ACL. There is no JarvisOS-held key to lose and none to rotate.
- This machine is shared. `s42-watchdog.service` is another operator's auto-recovery agent and it has root here; it can restart containers underneath you. Read its log before concluding that something you did caused what you are seeing.
- Its Coolify instance is self-contained — it manages only itself. Recovering it needs nothing from home1 or VIN, which is the entire point of the node.
Bring-up order
Dependencies first — starting out of order is a common second incident.- VPS1 / de-james-vps1 Node target
Readiness breakdown
- Documented runbook 20 pts
- No backup configured 50 pts
- No failover target configured 30 pts