The IT industry loves acronyms, almost as much as marketers love buzz words and trends. However, combining the two can lead to a rather muddled vat of terminology that obscures meaning, rather than adding clarity.
So, I’d like to help provide that clarity by boiling it all down and giving some perspective on the differences between backup, disaster recovery (DR), and Disaster Recovery as a Service (DRaaS), addressing how and when you should think about using these technologies.
A backup is simply a snapshot of your virtual machine, typically taken once a day after close of business. Backups are stored for a set period of time, with a retention policy such as "one a day for a month and one a month for a year." They are usually kept in a place that is separate from the primary system, like a basement, another data centre, or, in the case of Backup-as-a-Service, in the cloud. In my opinion, backups are great for legal and regulatory reasons, data protection and so on. However, they are not so great for business continuity.
Why do I say this? The most recent copy you have is from last night, which is usually stale as most business applications change significantly during a working day. Equally, to re-awaken a system from a backup takes time. First, you need a ready and working system, rebooted and running smoothly, and then you have to mount the backup. If the backup is (more safely) stored elsewhere, there is also the travel time of getting it to the primary location to factor in as well. The whole process is not renowned for being very fast.
So how does this compare to Disaster Recovery? Disaster Recovery (DR) systems replicate your data on an ongoing basis from a primary location to a secondary location. Usually, replication periods are measured in minutes, or even seconds, so the last copy of your data is fairly fresh. The secondary copy is held in stasis until it is re-animated via a failover trigger. The secondary location can be another data centre in another region, or even the cloud, in the case of DRaaS. In my opinion, DR is fantastic for business continuity because you can trigger a failover and be live at the secondary location in minutes. While some systems do enable an archiving-type function, most people agree that replication itself is not really the right tool for long-term data storage.
This is where I believe that Disaster-Recovery-as-a-Service (DRaaS) really comes into its own. If you don't have access to a secondary location, or you don't want to pay for it, DRaaS enables you to execute the same Disaster Recovery replication with the cloud as your secondary location. In this instance, the speeds involved are similarly fast—it takes minutes or seconds to recover systems. With some providers, you only pay for storage while the systems are replicating/in stasis, and once they failover, you'll pay for CPU and memory resources as you consume them. That said, you will want to vet the cloud as an operating environment for your workloads. Levels of security, customer support, compliance and transparency vary greatly from provider to provider.
It’s important to understand the needs of your business and choose a DRaaS provider, accordingly. DRaaS is a great way to achieve business continuity goals quickly and more cost-efficiently than owning a secondary data centre yourself.
Monica Brink, Director of Marketing EMEA, iland
Photo Credit: Olivier Le Moal/Shutterstock