The ongoing tug-of-war between flat IT budgets and escalating business demand for disaster recovery (DR) shows no signs of abating. This is driving IT organizations to become more creative as they attempt to devise new ways to get the most out of their IT budgets.
A major attraction to outsourcing of any kind is cost reduction. Building a DR capability has historically been an expensive proposition with significant investments in capital equipment such as servers and storage, bandwidth, and even secondary data centers. While technologies like virtualization can reduce some up-front costs, the perception of DR as an expensive insurance policy persists. But with newer and improved technology, outsourcing disaster recovery can potentially lower its costs.
Three factors when considering outsourcing disaster recovery
Outsourcing disaster recovery does not always guarantee lower costs. Three factors must be considered to determine whether outsourcing DR is appropriate for your environment:
- Requirements -- You must first understand the specific DR requirements from a business perspective. Whether insourcing or outsourcing, moving forward without a solid definition of business needs will likely result in overspending, dissatisfaction, or both.
- Options -- Make a realistic assessment of all available outsourcing options, including their actual capabilities and where they fit best. All disaster recovery outsourcing is not the same, and there are a variety of solutions designed to address varying needs.
- Existing capabilities and environment -- Understand existing skill sets -- or lack thereof -- as well as infrastructure, facilities and other organizational factors that may play a substantial role in tilting the scales either toward or away from outsourcing.
Disaster recovery outsourcing options and permutations
There are many disaster recovery outsourcing choices ranging from traditional hot-site services to basic hosting facilities. In addition, there are also a growing number of cloud-oriented choices designed to fill specific gaps in a DR strategy.
Traditional DR outsourcing specialists, including IBM Corp., SunGard and others, represent the category that perhaps is most likely to come to mind when one mentions the words disaster recovery and outsourcing in the same sentence. These organizations have comprehensive portfolios of services that range from contingency cold sites with the commitment to provide infrastructure on a contractual basis to hot sites that act as data replication targets and maintain servers at the ready. In addition, they also offer complementary services including personnel assistance, managed backup and recovery, staff business continuity (BC) support and more. All of this comes at a cost, and while these organizations, like others, are leveraging technologies such as virtualization to keep prices down, the more options and resources that are required the higher the bill.
Another option considered by some organizations is to simply rent or lease data center space from a hosting provider to serve as a DR facility and continue to manage it internally. While renting or leasing data centers avoids the build-out cost of a data center, they do little else to facilitate DR -- IT still owns and manages it. Still, a suitable data center is an increasingly expensive proposition, and renting the space may be all that an organization requires to enable their DR capability, particularly if they are already heavily virtualized. Also, many hosting providers are beginning to offer ancillary services that begin to approach capabilities offered by full-service providers.
Cloud disaster recovery
Cloud disaster recovery is sparking a lot of interest lately for data storage administrators. The cloud can offer several services that may be used to support DR. First, basic server and storage infrastructure services, such as those available through Amazon EC2, may be employed as an extension of existing data center resources, but targeted to address disaster recovery needs. Also, data can be replicated to cloud storage, and when needed, cloud servers can be activated.
From a cost perspective, both data storage and server resources are billed on a consumption basis, setting up the potential for avoiding costs of idle servers that can become unnecessary expenses for DR. On the other hand, there are a number of technical hurdles involved with integrating a cloud infrastructure with traditional infrastructure for DR, including migrating and replicating data, ensuring application functionality in the cloud, security and performance. Currently, this remains largely a "design-it-yourself" option.
Other cloud-based options include basic services such as remote data backup and storage that can fill an important niche need, but often do not solve the entire DR problem. But startup service provider Simply Continuous Inc. offers a solution to this problem. They focus specifically on the area of cloud-based disaster recovery and use virtualization and deduplication to replicate application images as well as data to their hosted recovery environment. Cloud-based DR is still in its infancy, but keep an eye out for more enhancements that use virtualization in combination with other technologies in the future.
Is disaster recovery outsourcing right for you?
The first step to determining whether outsourcing DR is viable and which option might be best begins with defining your individual requirements. The following questions are important to consider about before you choose a DR outsourcing option:
- What are your recovery point objectives (RPOs) and recovery time objectives (RTOs)?
- What are your DR performance expectations and requirements? And how much time does it take to operate a recovery site?
- Can you afford working in a degraded mode for more than a limited time?
To evaluate outsourcing, it's helpful to start by determining the cost and effort of providing the required disaster recovery capability in-house, and then weigh the various outsourcing options based on their ability to address defined requirements compared to internal delivery costs. Additional factors that will impact their viability include security requirements, bandwidth needs, performance impact, requirements for supplemental support and expertise and specific application mix and workload.
Whether building or outsourcing, the disaster recovery landscape is changing. Not so long ago, outsourcing DR largely meant third party facilities, servers and storage to host recovery tapes in a disaster. The options have expanded as newer technological innovations have evolved, providing organizations with a multitude of choices. With choices comes the challenge of selection, and the better your needs and capabilities are understood, the better choice you will make.
About this author: Jim Damoulakis is CTO of GlassHouse Technologies, the leading independent IT infrastructure services provider. He can be reached at firstname.lastname@example.org
This was first published in January 2010