Cloud disaster recovery (cloud DR) is a backup and restore strategy that involves storing and maintaining copies of electronic records in a cloud computing environment as a security measure. The goal of cloud DR is to provide an organization with a way to recover data and/or implement failover in the event of a man-made or natural catastrophe.
There are a number of benefits that make cloud disaster recovery appealing, including the variety of ways it can be implemented: in-house, partially in-house or purchased as a service. This flexibility allows smaller enterprises to implement robust disaster recovery plans that would otherwise have been impossible. Typically, cloud providers charge for storage on a pay-per-use model, based on capacity, bandwidth or seat. Because the provider is in charge of purchasing and maintaining its storage infrastructure, the customer doesn't have to spend money on additional hardware, network resources, data center space and the personnel required to support them.
In addition to cost, there are other important issues to consider before adopting cloud-based disaster recovery:
- Does the organization have the necessary bandwidth and network resources to move data fast enough between the primary site and the cloud?
- Can the organization encrypt data in flight as it leaves the data center?
Failover, failback keys to cloud recovery
Effective cloud disaster recovery provides continuity for services and the ability to fail over to a second site if there is a hardware or software failure of IT systems. Workloads are then failed back to their original locations when the crisis is resolved. Failover and failback can be automated. Organizations should run tests at regular intervals on isolated network segments that do not impact production data.
Organizations can choose to fail over data, entire applications or virtual machine (VM) images. When data is failed over, it is available from file services in the cloud. However, cloud recovery can take a long time if there is a great deal of data. Application-based data can be replicated to another application running in the cloud. Or an entire VM image, including data, can be replicated to the cloud and powered up and accessed if there is an on-premises failover.
Cloud service-level agreements
Service-level agreements (SLAs) hold cloud providers accountable and establish recourses and penalties if providers don't live up to their promises about cloud services.
SLAs can call for the provider to reimburse customers
with credits if there is a service outage or data cannot be recovered during a disaster. Customers can usually apply credits toward their cloud bill or a subscription to another service, but these credits seldom make up for the loss of business if cloud recovery is delayed. Customers should also study SLAs to help formulate an exit strategy for the service.
SLAs for cloud disaster recovery can include guarantees for uptime, recovery time objectives (
RTOs ) and recovery point objectives. For instance, an RTO can be from one hour up to 24 hours or even longer, depending on how important an application is to restore the business. The faster the guaranteed restore time, the more expensive the service costs.
Cloud disaster recovery providers, vendors
Because the cloud removes the need to maintain a second site, DR is considered a prime use case for the cloud. Disaster recovery requires failing applications over to the cloud and
failing back, so hundreds of vendors have sprung up to offer cloud DR services. The leading cloud DR as a service vendors include Axcient, Bluelock, IBM Resiliency Services, iland , Microsoft Azure Site Recovery and SunGard Availability Services.