Status. Proposed design only. Nothing in this vault is DG client approval. No application, service or test has been executed; every scenario is Not executed. Fees and dates are open.

Deployment, recovery and ownership

Vendor-neutral topology (architecture rev 3 p. 9). DG owns production accounts and funds hosting, identity, email and any later providers; vendors, region and identity are DG decision D08 (Q09).

flowchart LR
  SC[Version-controlled source + migrations] --> CI[CI: tests / build / scan] --> ST[Staging<br/>synthetic data] --> UAT[DG UAT + readiness<br/>written launch approval] --> PR[Production<br/>web + worker + DB] --> OP[Backup / restore<br/>monitoring / rollback]
AreaPlan / evidence
Environment separationLocal, staging, production have separate data and credentials. Staging uses synthetic or expressly approved sanitised examples. No live sensitive data until DG data policy is agreed.
Deployment and rollbackVersioned builds and migrations; reversible rollout where feasible; migration backup/forward-recovery plan; smoke checks; documented rollback decisions.
OperationsMonitor API errors, unauthorised attempts, worker lag, failed delivery and provider usage without private bodies in logs. DG names support and incident owners.
Recovery and availabilityDemonstrate restoration and record measured recovery. Balanced load/recovery targets need agreement. The proposal’s 99.5% monthly availability is a planning reference, not an SLA or a staging-proven result.
HandoverRepository, configuration inventory, deploy/rollback and restore runbooks, failed-job recovery, escalation routes, provider ownership. Retention/deletion policy is DG-approved.

Release sequence: staging evidence → DG UAT acceptance → separate written launch approval → smoke check and monitoring. Production is never deployed on silence.

Scenarios: T37, T38, T39, T40.