Case Study On Hybrid Disaster Recovery

Scroll down
Zacary Fettig
Zacary Fettig
Tech
  • Pine Cove Consulting MSP
  • Azure Solutions Architect Expert | MCSE: Core Infrastructure

Case Study On Hybrid Disaster Recovery

Overview

Evaluating disaster recovery (DR) strategies for a hybrid infrastructure environment by comparing three common approaches: replication-based recovery using Azure Site Recovery (Azure ASR), hypervisor level replication, and backup based recovery. The goal is to analyze tradeoffs between recovery time, data loss, cost, and operational complexity in order to determine which strategy is most appropriate for mid sized organizations.

Scenario

A regional credit union operates 12 branch locations connected to a central data center. The primary site hosts a VMware based environment consisting of three ESXi hosts backed by shared SAN storage. The SAN is configured with dual controllers and an Azure based witness to provide quorum for high availability and split-brain prevention. Critical workloads include core banking applications, authentication services, file services, and internal web applications.

The organization is looking for an option that can quickly restore critical banking workloads in the event of a primary site outage, ensuring continued operations across branch locations.  Key considerations include minimizing downtime, maintaining cost efficiency, and avoiding unnecessary operational complexity in the event of failover.

Preferred DR Strategy Approach

Based on the organization's requirements for rapid recovery, minimal data loss, and cost efficiency, Azure Site Recovery (ASR) is the most suitable disaster recovery solution. Unlike backup-based recovery, which can require hours to restore services, or hypervisor-level replication, which often requires a secondary on-premises environment, Azure ASR provides continuous replication of virtual machines to Azure, enabling failover without maintaining duplicate infrastructure. With a recovery point objective (RPO) around three to five minutes and a recovery time objective (RTO) of approximately 13 minutes in my testing of ASR, Azure ASR minimizes both data loss and service interruption while remaining cost-effective for a mid-sized organization. Additionally, the organization's existing Microsoft-centered environment makes ASR a natural fit, allowing the credit union to leverage its existing Microsoft technologies, management tools, and administrative expertise rather than introducing an entirely separate DR platform. This allows the credit union to quickly restore critical banking applications, authentication services, file services, and internal web applications during a site-wide outage, ensuring continued operations across all branch locations.

Alternative DR Solutions

Although Azure Site Recovery provides the best balance of recovery speed, data protection, and cost for the organization's requirements, other approaches including the ones below were considered.

Veeam Backup and Replication

Veeam provides two relevant approaches for disaster recovery: backup-based recovery and VM replication. Veeam backup based recovery can provide strong protection against data loss and ransomware by maintaining multiple recovery points, but recovering the environment in Azure requires restoring the virtual machines before they can become operational. This additional restoration process can increase the RTO compared with Azure Site Recovery's continuously replicated workloads. Veeam replication provides faster recovery because replica virtual machines are maintained at a secondary target and can be failed over rather than fully restored from a backup. However, a traditional Veeam replication architecture requires an available target virtualization environment to continuously maintain the replicas. For the credit union, maintaining a secondary VMware cluster would introduce additional hardware, licensing, storage, and operational costs. While Veeam is an effective solution, its infrastructure requirements make it less cost effective than ASR for a disaster recovery environment that only needs to operate when the primary on premises site is unavailable.

Cloud Based Hypervisor Replication

Azure VMware Solution (AVS) provides another alternative by extending VMware infrastructure into Azure. AVS would allow the organization to maintain a VMware environment in Azure and recover workloads using the same VMware technologies used at the primary site. However, AVS requires dedicated Azure VMware infrastructure to remain deployed even when there is no failover event. This creates substantially higher ongoing costs compared with ASR, where replicated workloads can remain in Azure storage without requiring the Azure VMs to run until a failover occurs. This same cost consideration applies to other DR solutions that require continuously running target infrastructure, as the organization must maintain and pay for the recovery environment even when no disaster has occurred. AVS would be a better fit if the organization intended to maintain a permanent VMware environment in Azure, rather than using Azure exclusively as an emergency disaster recovery site.

Traditional Secondary VMware Site

This approach can provide solid stability in the event of a failover, but it requires additional servers, storage, networking, licensing, power, and facility costs. Azure Site Recovery eliminates the credit unions need to maintain a full secondary datacenter for cost savings and decreasing complexity of maintaining another full on prem environment.

© 2023 All Rights Reserved.
Email: zacaryfettig@gmail.com
Write me a message
Write me a message

    * I promise the confidentiality of your personal information