Cloud disaster recovery services vs. traditional DR: A weigh-in

Do cloud disaster recovery services’ benefits outweigh those of traditional DR? An expert weighs the viability of these disaster recovery approaches.

Of late, many organizations have started to use the public cloud to either host non-critical applications, or for storage. Some have gone a step further, and moved part of their critical applications to the cloud. Very soon, organizations could be seriously considering service providers which offer cloud disaster recovery (DR) services.

The cloud disaster recovery service concept promises several benefits over traditional disaster recovery service, including lower cost of operations and faster recovery times. While it may not be a replacement for traditional disaster recovery setups, there are several scenarios where a cloud DR service works better than traditional DR.

Cloud disaster recovery service over traditional DR

  • Cost is the key factor for choosing a cloud disaster recovery service. A secondary physical DR site means investments in additional data center space, connectivity and servers. It also leads to additional operational costs—power and cooling, site maintenance, and manpower requirements.

    A cloud-based disaster recovery service provides virtual machine snapshots of physical or virtual servers at the primary data center. The organization pays for storing the snapshots, application data in a suspended state, and replication of data from primary to the secondary (cloud DR) site for data synchronization. It pays for the infrastructure-as-a-service feature only in case of a disaster, wherein virtual machines (snapshots of primary servers) need to be brought online as a substitute for the primary site.
  • With cloud disaster recovery services, the DR site can be brought online within seconds or minutes—as opposed to a physical DR site which could take minutes (if not hours). Booting a physical machine takes at least a minute or more, while a virtual machine instance can be up and running within seconds. Typically, a physical DR site operates only during data replication, or in the event of an actual disaster. The time taken to make a DR site live will take more time, in comparison to a cloud DR. In addition, data loss is directly related to downtime. A cloud DR site that boots up within a few seconds translates to data loss of just that timeframe.
  • In case connectivity is not available with the physical DR setup, manual operations may be required to start the site’s operations. However, a cloud based disaster recovery service can still be triggered from a laptop using a wireless internet connection.

Traditional disaster recovery over cloud DR

  • Typically, a company’s physical DR site is located at the nearest feasible geographical location. On the contrary, a cloud service provider may base data center(s) in a different continent. This can cause latency issues for a company whose virtual machine (VM) instances are stored in that data center. A cloud disaster recovery service may not be suitable for a company which has critical applications that demand high response times and low latency.
  • With a cloud disaster recovery service, the organization needs to ensure that its applications are compatible with public cloud infrastructure. For instance, certain applications may demand a particular environment that may not be available with the cloud service provider. While cloud services might support environments with common application customization, it will not support strongly customized applications. A cloud DR service is not feasible for environments with such application stacks.

    Choosing to go with a cloud disaster recovery service will be governed purely by the business imperative. If an organization has critical applications that should be available within minutes of downtime, it can consider cloud based DR. However, if an organization has restrictions on where the data resides (geographic location), latency issues, or custom applications, it should opt for physical DR.

About the author: Simran Gambhir is an independent IT consultant, with expertise in areas such as strategy, solutions architecture, process and execution planning. He has in the past held the position of CTO at organizations such as Praxeva India Services, Loyalty Pacific, and News Interactive (Australia).

(As told to Harshal Kallyanpur.)

Read more on Disaster recovery

Data Center
Data Management