|By Nicos Vekiarides||
|October 15, 2012 04:00 AM EDT||
While many new technologies claim to revolutionize the practice of disaster recovery (DR) for IT environments, few have significantly altered the economics and logistics of building and maintaining a secondary IT site. For most organizations, a secondary site housing server and storage infrastructure has remained the only recovery path for their business from a primary site disaster, failure or outage.
While it might seem preferable to avoid the expenditure of a secondary IT site altogether, what drives investment in DR infrastructures is a set of recovery time objectives (RTOs) and recovery point objectives (RPOs) that respectively determine the maximum allowable downtime and data loss an organization is willing to sustain. While certain organizations may require both of these objectives to be near zero (i.e. instant recovery, no data loss), other organizations may be able to withstand minutes or even hours of downtime. Understanding these objectives is one of the fundamental tenets for DR planning and investment.
Who shot DR?
Cloud computing and storage, offering on-demand activation of IT resources and an efficient pay-as-you-go consumption model, represent the first real transformation and rethink of disaster recovery deployments. The very nature of cloud-based IT — cloud meaning service-based offerings that requires near-zero maintenance rather than relabeled and repackaged traditional infrastructure — eliminates capital expenses on dedicated infrastructure with no need for significant investment until a true disaster occurs. This translates to peace of mind for IT in areas prone to natural disasters, without the expense of a secondary site. Attractive as that may seem, a reasonable question may be whether cloud DR solutions can meet the RPOs and RTOs that businesses require.
Meeting recovery objectives entails regaining access to both data and applications after a disaster or outage. The recovery process involves accessing data and spinning up applications either on-premise or, preferably, in a cloud environment. Cloud storage gateway and cloud-integrated storage technologies have simplified storing data or live copies of data in the cloud, enabling data access from virtually anywhere, including in the cloud. On the other hand, on-boarding applications to cloud for DR calls for interoperability of the the local compute environment and cloud compute environment so that applications can recover gracefully. While this is feasible, we will examine some of the considerations in standing up applications in a cloud environment in the next part of this series.
When it comes to meeting recovery objectives, keep in mind there are some limitations when bringing up a disaster recovery environment on-demand. Unlike the case of dedicated disaster recovery, a cloud environment necessarily implies a “spin-up” delay to activate the recovery environment in the cloud. For organizations requiring zero down time, cloud may not be an option or may not even be cost-effective when attempting to meet the most stringent of RTOs. However, for organizations seeking disaster recovery in minutes or hours versus days, with recovery times significantly better than tape backup and restore, cloud can offer both an improvement in recovery objectives and a potentially huge savings in costs.
Bottom-line? It behooves nearly any organization seeking to reduce disaster recovery costs and improve availability of their IT environment to consider cloud-based disaster recovery options.
Look for part II of this series which will discuss some of the more intricate considerations of application and data recovery using the cloud.