The adoption of flexible and cutting-edge solutions to achieve greater stability, continuity and resilience of the main application workloads that support the company business is an important goal to pursue. Azure VMware Solution (AVS) is the service designed, made and supported by Microsoft and approved by VMware, which allows customers to use physical VMware vSphere clusters hosted in Azure. This article describes the main Azure VMware Solution adoption scenarios to meet disaster recovery needs.
The Azure VMware solution can be contemplated to address different scenarios, including the implementation of disaster recovery and business continuity plans (BCDR). The following diagram is useful for guiding the possible choices in this area at a high level:
Figure 1 - Diagram to guide choices in the BCDR area
Disaster Recovery Strategies Considerations
First of all, it helps to align business requirements with RPO, RTO and with the availability of IT resources. An effective Disaster Recovery plan must be designed to achieve these goals by adopting the most appropriate technologies. About this, the adoption of native BCDR solutions for applications can be evaluated, for example SQL Always On availability group or SAP HANA System Replication (HSR), or non-native solutions such as VMware Site Recovery Manager (SRM) and Azure Site Recovery.
Azure VMware solution should be evaluated if, in enterprise environments, there are particularly stringent requirements in terms of RPO and RTO. Otherwise, you can use Azure Site Recovery or rely on system recovery using the Data Protection solution you are using.
Disaster Recovery Solutions using Azure VMware Solution
To implement disaster recovery plans using Azure VMware Solution, it is possible to adopt native VMware solutions or third-party solutions.
VMware Site Recovery Manager (SRM)
VMware Site Recovery Manager is an automation solution, which integrates with underlying replication technology, able to offer:
- Recovery test without service interruptions
- Workflow able to carry out the orchestration of DR plans in an automated way
- Automatic reset of network and security settings (integration with VMware NSX)
The solution offers the possibility to insure, in a simple and reliable way, restore and move virtual machines between multiple VMware sites with little or no downtime.
Site Recovery Manager requires one of the following replication technologies to orchestrate virtual machine recovery operations:
- VMware vSphere Replication: replication focused on VMs and based on the hypervisor. It is the solution natively integrated with Site Recovery Manager and included in most versions of vSphere.
- Third party solutions: Site Recovery Manager uses plug-in SRA (Storage Replication Adapter) developed by storage partners for integration with third-party systems.
Site Recovery Manager (SRM) for Azure VMware Solution (AVS) is able to automate and orchestrate failover and failback processes in the following Disaster Recovery scenarios:
- On-premise VMware to Azure VMware Solution private cloud disaster recovery.
- Primary Azure VMware Solution to a secondary disaster recovery Azure VMware Solution private cloud.
Furthermore, thanks to the possibility of carrying out failover tests, without generating interruptions on the production environment, it is possible to periodically guarantee the achievement of the objectives, related to the recovery time, required for disaster recovery plans. In this scenario SRM is licensed and supported directly by VMware.
For further details you can consult this Microsoft's document.
VMware HCX Disaster Recovery (DR)
Although the VMware HCX solution can be used with the Azure VMware Solution private cloud as a recovery target or destination, this possibility should only be considered in particular cases. In fact,, the adoption of this solution is not recommended for large environments, as the orchestration of activities during disaster recovery is totally manual. Currently the AVS solution does not have runbooks or features that can support failover operations during a disaster recovery. For enterprise DR scenarios it is therefore recommended to use the VMware Site Recovery Manager solution (SRM) or the third-party solutions described in the following paragraph. For more details about DR with VMware HCX you can consult this Microsoft's document.
Third party solutions
Microsoft, in order to guarantee its customers the opportunity to make the most of the investments made in skills and technologies, has partnered with some of the industry's leading partners to ensure integration and support. Among the main third-party solutions that can be used we find:
The adoption of some of these solutions, like JetStream, can also be advantageous in terms of cost, as Azure Blob Storage is used to keep copies of virtual machines and related data.
Figure 2 - DR operations with solutions such as JetStream
The data written by the virtual machines locally is replicated directly to an Azure Blob Storage and the resources of the DR infrastructure are not needed until a disaster occurs (or when a test is required) which prompts you to initiate the disaster recovery procedure. This approach implies that costs are minimized, with a consequent reduction in infrastructure costs compared to alternative DR solutions.
When an event occurs that involves the activation of the DR plan, the application and configuration data stored in the Blob Storage come quickly “rehydrated” in a vSphere cluster activated in AVS, allowing applications to resume operation in AVS.
Conclusions
When you need to develop a disaster recovery solution, each company has specific requirements to be respected and unique preferences on where to place workloads and how to manage the emergency plan. Azure VMware Solution (AVS) can be used flexibly according to different implementation models, easily integrating with both native VMware solutions and third-party solutions. This allows, to organizations at the enterprise level, to be able to rely on Azure VMware Solution for their DR scenarios, with the guarantee of obtaining effective protection and respecting corporate objectives in the Disaster Recovery field.
