Decommission

Destructive — follow the order exactly

Owner: automation-guild · Auto-drafted from ansible-decommission.

Preconditions

• Decommission approved (CHG____ / REQ____)
• Dependencies checked — nothing still consumes this resource
• Final backup captured per retention policy
• Stakeholders notified

Order of operations

1. Drain / disable traffic
2. Final backup to geo-redundant storage (retain per policy — 90d for SQL)
3. ansible-playbook decommission.yml -i <inventory> --limit <group>
4. Remove from monitoring (SolarWinds / Azure Monitor) and CMDB
5. Release IPs / DNS / licenses, close billing tag

Validation

• Absent from CMDB
• No orphaned monitoring alerts
• Billing tag closed

Backout

• Restore from the step-2 backup if aborted mid-run.

Related

• Provisioning