In the past, a “disaster recovery” plan was the nightly tape backup. For extra protection, you might store the tape off site. But in the past, customers didn’t shop online at 2 am or expect 24-hour customer service and overnight delivery. Today, disaster recovery (DR) means keeping business systems running around the clock, without interruption or loss of data.
A good DR strategy with fast failover and recovery can help a business survive a ransomware attack, power outage, server failure, or natural disaster. It’s even possible to have almost zero downtime, with immediate recovery of operations and no data loss.
However, designing an effective and affordable DR solution requires research and planning. IT environments today are more complex than they were 15 or 20 years ago, and those complexities create challenges in DR.
Dale Levesque, director of product management for TierPoint, spoke about DR and disaster recovery as a service (DRaaS) during the Enabling Next-Generation IT Summit. Levesque described four key DR challenges and how DRaaS can help.
Key disaster recovery challenges
Servers vs applications
DR is about more than backing up servers and data. Today’s applications depend on multiple components—web servers, app servers, databases—sharing processes and data. An ecommerce system, for example, sends and receives data with inventory applications, finance applications, identity and access management, and logistics. To backup and restore an ecommerce system, a DR strategy has to encompass all of its related servers, data, processes, and workloads.
Too often, IT organizations neglect to thoroughly inventory their IT environments, including platforms, applications, storage, network hardware and services, development servers, app servers, and so forth. Without a complete understanding of the environment, IT can’t map out the dependencies between the components. If server A is down, what does that mean for servers B and C? A good DR strategy requires knowing your environment.
When backing up a complex system, it’s critical to synchronize the schedule for all of that system’s components. An ecommerce system, for example, includes web servers, app servers, storage, and databases. If those aren’t copied at the same times, the completed backup may not be recoverable or lack the latest transactions. One solution to this problem is to use virtual protection groups, which link all of the virtual machines that support a service or application. A protection group might include any combination of application, web, and file servers and database clusters. Virtual protection groups ensure everything is synchronized when copied.
Knowing your RPO and RTO requirements
Not every application needs to be brought back online immediately. Some can wait a few hours, or even a day or two, while others need to be back up immediately, with no loss of data.
Your recovery time objective (RTO) and recovery point objective (RPO) will determine what type of DR solution you need.
RPO—recovery point objective—indicates how much data you can afford to lose. An RPO of seconds might be appropriate for a patient monitoring application, whereas an RPO of an hour or more is possible for less critical applications such as administration.
RTO is the maximum amount of downtime that is acceptable. An ecommerce application at an online retailer requires an RTO of seconds or less. Any longer could force customers to shop elsewhere. Productivity applications, however, could potentially be down a day or two without jeopardizing the business.
The most mission critical applications, which require the lowest RPO and RTO, are typically infrastructure applications like VPNs and network software, followed by customer-facing or revenue generating applications, such as ecommerce or web servers. The least critical applications are likely to be archival file servers and niche applications that few employees use.
How DRaaS can help address disaster recovery challenges
A DRaaS provider offers access to third-party experts who can consult on a DR strategy, inventory IT infrastructure, and help to select the RPO and RTO of different applications. DRaaS also frees the IT department from the need to maintain and update the DR software and hardware, because it’s managed in the cloud by the provider.
Cloud DRaaS also supports faster RTO and lower RPO because backups and restores can be done directly to and from the cloud. With no intermediary tape backup, the process can be done much more quickly.
As more organizations adopt complex hybrid IT environments, the need for DR grows as well. Fortunately, DRaaS providers can help take the burden of DR planning and management, and guarantee consistent, reliable, and rapid backup and recovery.
Watch Dale Levesque’s full presentation: Enabling Next-Generation IT Summit.