Disaster recovery plans are a series of steps you take to repair damage and carry on. But a DR plan won’t work when you have no time to rebuild, says a recent report by Forrester Research Inc.
The goal today is 100 per cent uptime. But how do you get there? According to the analyst firm, IT infrastructure has to be constructed from the ground up to be resilient. In a report titled, “Move beyond disaster recovery and prepare for business technology resiliency,” Forrester decries what it says is an outdated concept of leaving DR to storage professionals and argues it should be replaced by a comprehensive, company-wide strategy: business technology resiliency.
While certain businesses have had to maintain round-the-clock uptime for years, often at great cost, new technologies such as cloud computing have made it possible to achieve it more cheaply, the report says. Yet many companies are relying on legacy systems for even their most important data, with nearly 40 per cent of companies still using tape to recover mission-critical applications, it adds.
Aside from being potentially dangerous, relying on tape is also inefficient given the recent advances in virtualization and cloud technology, the report says.
But even when companies use virtualized or cloud environments to replicate data and build more robust infrastructure, it’s often the software that lags behind:
“Oftentimes the biggest barriers to adopting active-active or load-balanced architectures are the applications themselves — developers did not code them to sustain this type of approach. For those companies that do have applications that IT can run in an active-active approach, there is the additional barrier of high licensing costs for running two copies of an application.”
Using the analogy of one jet engine on a plane failing and another immediately picking up the slack, Forrester points out that applications can be made more resilient by building the expectation of disaster occurring into their designs.
This can mean not only having in-house developers code this way, but also encouraging third-party vendors to produce software with more built in availability and resiliency (the report notes that Microsoft and Oracle were some of the first to make steps in this direction, with VMware, Research in Motion Ltd. and SolarWinds Inc. following suit in recent years).
Businesses can also tell storage providers they won’t pay premiums for replication, the report suggests. In any case, this is an extra cost some of these vendors are already cutting out as bandwidth prices come down.
In a broader sense, to achieve maximum uptime, companies should aim for service continuity as opposed to infrastructure continuity, Forrester says. In the past, they have focused too much on the latter and end up missing the bigger picture: availability on the customer side.
“Ultimately, the companies that have been most successful with embedding resiliency have been the ones who embrace the concept of DevOps with improved collaboration, shared responsibilities for failures, and constant visibility between app dev, infrastructure engineering, and operations,” the report adds.
Collaboration is crucial for business technology resiliency, Forrester stresses, with some companies having overlapping plans for disaster recovery, business continuity and security response that aren’t shared across the business. Much better, it says, to involve everyone in one comprehensive plan.