Not long ago, disaster recovery was something that only the largest corporations attempted to implement because of its staggering costs and extreme complexity. But in recent years, cloud service providers such as Amazon and Microsoft have brought DR capabilities to the masses with such products as AWS CloudEndure and Azure Site Recovery. Today, nearly any organization can replicate its mission-critical servers to a public cloud and then fail over to a cloud-based replica in the event of a disaster.
AWS purchased startup CloudEndure in January 2019, adding to its portfolio continuous data replication functionality that enables DR in the cloud.
In 2014, Microsoft rebranded Hyper-V Recovery Manager to Azure Site Recovery. This service continues to be one of the most popular and proven tools to provide automated protection and DR in the cloud.
Examine these two cloud-based DR heavyweights and compare the workloads each supports as well as their architectures, features, integrations and pricing models.
AWS CloudEndure vs. Azure Site Recovery workload support
CloudEndure is built to protect database workloads. The service works with Oracle, MySQL and Microsoft SQL Server, as well as with select enterprise applications, including SAP, Microsoft Exchange Server and Microsoft Internet Information Services. Organizations can also use the service to protect other types of virtualized workloads by providing replication services for VMs.
Azure Site Recovery provides application-aware replication for Microsoft server applications such as Exchange Server, SQL Server, SharePoint Server and Dynamics. It also works with Active Directory and several third-party vendors to provide support for common business applications, including SAP, Oracle, Red Hat and IBM.
Comparing AWS CloudEndure vs. Azure Site Recovery architecture
As previously noted, CloudEndure protects workloads at the VM level. To enable this protection, administrators must deploy a CloudEndure agent to the VMs they wish to protect. Once deployed, this agent replicates the VM's OS, system state, files, applications and databases to a staging area within the AWS cloud. This staging area is based on the use of low-cost Elastic Compute Cloud (EC2) VM instances, with the actual data residing in Elastic Block Store volumes. The replication process is ongoing, allowing for recovery point objectives (RPOs) of seconds, with recovery time objectives (RTOs) measured in minutes.
If a failover becomes necessary, then the EC2 instances residing in the staging area are converted into production-grade instances, suitable for hosting the protected workloads. The service also allows for a nondisruptive failback.
Although the architecture of Microsoft Azure Site Recovery is similar to Amazon's CloudEndure, it can differ slightly, depending on the source that's being replicated to the Azure cloud. In the case of a Microsoft Hyper-V host or host cluster, for instance, an agent known as the Microsoft Azure Site Recovery Services agent handles the data replication process. In addition to the agent, Azure Site Recovery also uses an Azure Site Recovery Provider, which is also installed in the primary site, the Hyper-V host or host cluster. This provides orchestration services during a failover.
Key features of CloudEndure and Azure Site Recovery
AWS CloudEndure uses a continuous data protection engine to provide asynchronous, block-level replication of source servers into a staging area. The staging area acts as a low-cost, cloud-based repository from which VM instances can be spun up in the event of a failover or a failover test.
While the staging area is primarily designed as a tool for enabling cloud-based DR, it's also able to provide granular, point-in-time recovery. This provides a layer of protection against ransomware attacks or accidental data deletion or modification.
CloudEndure's built-in orchestration engine automates the VM conversion process that's required during a failover or failback operation. The orchestration engine is built for scalability and, if necessary, the fail over or fail back of thousands of VM instances.
CloudEndure allows for nondisruptive drills, which enables an organization to test its DR resources without disrupting the production environment in the process.
Azure Site Recovery's feature set is similar to CloudEndure. Like CloudEndure, for instance, Azure Site Recovery is based on the use of data replication. It asynchronously replicates data every 30 seconds, which translates to RPOs of less than a minute. Admins can also use Azure Traffic Manager to reduce RTOs by prioritizing recovery traffic.
Azure Site Recovery also allows for nondisruptive testing. Admins can perform failover drills and tests without impacting the organization's production environment.
Microsoft Azure Site Recovery supports the use of customized recovery plans. This enables admins to control the sequence in which VMs are recovered, which can make the process of recovering multi-tier applications easier.
Azure Site Recovery also uses orchestration to automate the failover and failback processes. It also works with other Azure services and components to allow for automated IP address reservations, load balancing and network switchovers.
CloudEndure vs. Azure Site Recovery integrations
AWS CloudEndure works with both Linux and Windows VMs hosted on hypervisors, including VMware, Hyper-V and KVM. CloudEndure also supports workloads running on physical servers as well as cloud-based workloads running in AWS, Azure, Google Cloud Platform and other environments. Amazon publishes a full list of supported applications, hardware configurations, OSes and source infrastructures.
Azure Site Recovery replicates physical and VMs from a primary site to a secondary site within the Azure cloud. These physical or VMs can be running either Windows or Linux OSes. VMs running on premises can be hosted on VMware, Hyper-V or Azure Stack hosts.
Azure Site Recovery also supports replicating cloud-based VMs. An admin can replicate an Azure VM from one Azure Region to another. AWS EC2 instances support replication, as long as they're running Windows. Microsoft provides a full list of resources that it supports for use with Azure Site Recovery.
How Amazon's and Microsoft's pricing models compare
Amazon charges a flat hourly fee for its CloudEndure service. This fee is billed on a per-source server, or protected-host server, basis. The fee of $0.028 per server, per hour covers the cost of replicating data to the cloud, point-in-time recovery services, DR drills and orchestration for automated failover and failback. There's an additional fee for resources that are fully provisioned as a result of a DR drill or DR operation. Organizations can use Amazon's TCO calculator to estimate the total cost of using CloudEndure.
Microsoft bases Azure Site Recovery pricing on the total number of instances it protects. The service is free for the first 31 days, and then costs $16 per instance, per month for recovery to customer-owned sites and $25 per instance, per month for recovery to Azure. Note that pricing and available options vary by region. Microsoft charges additional fees for related items such as storage capacity, storage transactions, data egress and the compute capacity used during DR operations and drills. Microsoft provides a cost calculator to help organizations estimate the total cost of using Azure Site Recovery.
AWS CloudEndure vs. Azure Site Recovery support and training
Amazon offers free training for CloudEndure, and technical support is available through an AWS Business or Enterprise Support agreement.
Microsoft offers several different support options for its Azure cloud. Organizations can receive Azure Site Recovery support through Microsoft's Basic, Developer, Standard and Professional Direct support options. Microsoft provides comprehensive online documentation, including various tutorials and how-to videos.