TELEICU/Bulk migration playbook

playbookteleicuCARE 3.0+

Process for migrating the existing 200+ nodes from the current state to the target architecture in controlled waves.

Migration waves

{{#mermaid:flowchart LR
    W0["Wave 0<br/>Staging (test VMs)"]
    W1["Wave 1<br/>Canary (1-2 sites)"]
    W2["Wave 2<br/>1 state (pilot)"]
    W3["Wave 3<br/>3 states"]
    W4["Wave 4<br/>Remaining 3 states"]
    W5["Wave 5<br/>Validation & cleanup"]

    W0 --> W1 --> W2 --> W3 --> W4 --> W5
}}

Per-wave process

Before wave

- [ ] Verify staging passed all gating criteria - [ ] Confirm rollback procedure tested - [ ] Notify hub admins in the target region - [ ] Schedule maintenance window

During wave

- [ ] For each node: execute Site onboarding flow - [ ] Verify node health before moving to the next - [ ] Track progress in TeleicuSites - [ ] Pause wave if > 1 node fails

After wave

- [ ] Run full fleet health check on migrated nodes - [ ] Review VictoriaMetrics for anomalies - [ ] Collect hub admin feedback - [ ] Gate decision for next wave

Rollback during migration

If a wave reveals issues:

Stop migration — do not proceed to next wave
Roll back the affected wave's nodes (Rollback strategy)
Fix the issue in staging
Re-run the wave

Migration states

Tracked in the TeleicuSites table:

| State | Meaning | |---|---| | not-started | Running original deployment | | ansible | Partially migrated via Ansible path | | nixos | Partially migrated via NixOS path | | done | Fully migrated to target architecture |