Skip to content

DR Failover to Cloud

Executive summary

This showcase demonstrates a practical disaster recovery (DR) pattern where:

  • Primary workloads run on-prem in a Proxmox / Kubernetes environment.
  • A warm standby environment exists in the cloud.
  • Controlled DNS cutover is used to switch traffic during an incident.
  • Evidence is captured for both failover and failback.

The goal is to show a realistic, cost-aware DR approach suitable for SMEs and larger environments.

Case study – how this was used in practice

  • Context: Hybrid lab environment used to rehearse DR scenarios and support portfolio evidence.
  • Challenge: Show a DR strategy that is realistic, testable and not purely theoretical.
  • Approach: Built a minimal viable DR pattern: warm standby in cloud, DNS-based cutover, and explicit runbooks for failover/failback.
  • Outcome: Clear, repeatable DR steps with artefacts suitable for auditors, assessors and engineering leaders.

Related decisions (for example):

Demo

Video walkthrough

  • Video: https://www.youtube.com/watch?v=DR_FAILOVER_DEMO

The demo walks through:

  1. Normal operation with on-prem environment serving traffic.
  2. Simulated incident impacting on-prem services.
  3. DNS cutover to the cloud environment.
  4. Validation of application health and data paths in the cloud.

Screenshots

![DR architecture overview](./diagrams/dr-architecture-overview.png)
![DNS cutover screenshot](./screenshots/dns-cutover.png)

Architecture

  • High-level diagram:
![DR high-level architecture](./diagrams/architecture-overview.png)
  • Key components:
  • On-prem: Proxmox / Kubernetes cluster hosting primary workloads.
  • Cloud: Minimal standby environment sized for DR, not steady-state peak.
  • Control plane: DNS provider (and optionally load balancer) used to switch traffic.
  • State: Either replicated or re-hydrated from backups, depending on scenario.

Optional detailed diagrams:

Implementation highlights

  • Warm standby rather than always-on multi-region active/active to control cost.
  • DNS-driven cutover with clear TTL and rollback considerations.
  • Runbooks for both failover and failback, with evidence captured for each exercise.
  • Designed to be demonstrable in a home lab while mapping to real-world DR expectations.

Assets and source

  • GitHub folder for this showcase:
    https://github.com/hybridops-studio/hybridops-studio/tree/main/showcases/dr-failover-to-cloud

  • Infra and scripts:

  • infra/terraform/ – DR infrastructure definitions.
  • core/ansible/ – configuration for DR roles.

  • ./scripts/ – helper scripts used during DR drills.

  • Evidence:

  • ./evidence/ – screenshots, logs and exported dashboards from DR test runs.

Academy track (if applicable)

In the Academy, this scenario can be turned into a guided DR exercise where learners:

  • Execute the failover and validate application health.
  • Perform a controlled failback and verify state consistency.
  • Capture and review evidence to support an internal DR review.

Role-based lens (optional)

  • Platform Engineer / SRE: focuses on reliability, observability and safe procedures.
  • Infrastructure Engineer: sees practical replication and failover mechanics.
  • Engineering Manager / Hiring Manager: sees a realistic DR approach rather than a purely theoretical slide.

Back to showcase catalogue