Operations

Task-oriented index into the operational runbooks for each machine and each cross-cutting concern.

Active incident: arkk RAID recovery

The arkk RAID5 array is degraded (1 of 5 disks failed) and mounted read-only while a selective rescue backup runs to pyrite.

1. 2026 incident notes — current state, degraded-start procedure, resilver steps (deferred until backup verifies)
2. Backup triage — keep/drop plan, thinning policy, run procedure
3. General RAID recovery runbook

Machine runbooks

gatekeeper (VPS ingress and public services)

• Machine overview
• Service and port inventory
• Infrastructure layout
• Backup operations
• Tailnet / Headscale setup

pyrite (workstation, compute, and backup target host)

• Machine overview
• Platform baseline
• Boot and startup
• Remote access
• Troubleshooting
• Services on pyrite

arkk (NAS and storage)

• Machine overview — see active incident above

Cross-cutting

• Backup strategy
• Network documentation
• Secrets management
• Docker images

History

• Repo consolidation record (2026) — migration of admin/ and vps-infrastructure/ into this repo