Long gone are the days of simply throwing more storage at a problem. Instead, a more strategic approach that calls for the optimized, cost-efficient deployment of storage products and services has found its way to the forefront.
Increasing competitive demands, stringent compliance requirements and non-existent IT budgets are forcing organizations to realize that by combining their internal assets with judicious technology acquisitions, they can take control of their storage infrastructures, better enabling them to meet the new business demands placed upon information.
Information lifecycle management (ILM) — a strategy for proactive management of information which provides the ability to efficiently align information infrastructures with business needs based on information’s changing value — is a key enabler of these enhanced infrastructures. Tiered storage is the technological cornerstone of ILM.
With ILM tiered storage, users can match applications to the appropriate infrastructure for required service levels. For example, a mission-critical application with a highly transactional Oracle database would be supported by tier-one storage area network (SAN) storage, whereas file-and-print applications might be supported by network attached storage (NAS). When tiered storage architectures are associated with business service level requirements and optimized applications alignment, the result for businesses is a clear understanding of their platform requirements and a solid opportunity to decrease their total cost of operations (TCO).
Finding the Benefits Looking at information management from a practical perspective, the number of organizations using their tier-one primary storage for low priority applications such as file and print services is remarkable. Realigning these applications can yield enormous benefits to these organizations.
For example, consider the cost-avoidance savings that companies can suddenly realize when they are able to cut back new storage technology acquisitions because they have achieved efficiencies through application realignment.
Of course, realizing these significant benefits means overcoming the challenges most organizations face. The lack of cross-functional cooperation between IT departments and business users is well known. On one hand, business users have to understand their storage requirements. On the other hand, IT leaders have to satisfy those requirements with the right combination of products and services in a way that allows them to monitor and measure their performance.
Determining appropriate chargeback schemes driven by approved service level agreements (SLAs) puts even more pressure on these already dysfunctional relationships, and meeting these demands requires pooling storage architectures, which can cause additional turf wars. Balancing these conflicting inter-group issues with technology considerations is a delicate process, but an essential one that usually falls upon IT organizations.
Steps to Tiered Storage
The effort involved in defining and implementing a tiered storage architecture includes several key success factors, usually tackled in the order below.
1. Creating a storage services catalogue. Establishing a storage service catalogue is a critical aspect of implementing a tiered storage architecture because it provides a fundamental methodology for aligning the needs of business users with the resources of IT. A storage service catalogue includes the four storage service domains for which IT is responsible to deliver to business users:
• primary storage, which hosts active application data supporting normal business processes; • archival management, which handles the policy-based archive of application data;
• operational recovery which retrieves data in cases of server/application-related data failures;
• and disaster recovery which provides services for recovery of systems/data in the event of catastrophic data centre failure.
The first step in tiering is to determine which of the four service domains business users require. In some instances, the business only wants to tackle archive services supporting, for example, a compliance initiative. At other times, differentiating service levels for operational and disaster recovery is necessary for improving information protection levels. In the best circumstances, all four domains are treated in order to optimize the effort.
Once business and IT confirm the service domains to be included in the storage service catalogue, they must agree on specific service level measurements that will be delivered for each domain. For example, primary storage may include guaranteed performance and availability and, in some cases, volume sharing. More granular storage service levels are also established internally by IT so they may closely measure their performance in delivering against their business SLA. Most importantly, the service catalogue is a way for business and IT to agree on what services are important to the business and how these services will be measured.
2. Developing technical and operational support requirements. Once business needs have been validated via the storage services catalogue, the next step is to develop technical and operational support requirements. This analysis involves identifying the class or classes of technologies — but not the specific vendors — necessary to deliver the service levels, and matching that list against existing customer resources. By identifying the gap between current IT assets versus what is needed, enterprises are able to identify the cost of filling those gaps and estimate of the overall cost of delivering the service catalogue.
Inherent in the development of these requirements is the understanding that even the best technology will fail if it is not supported by the right operational capabilities. For that reason it’s also important to identify key operations capabilities such as people, process and management technology, to leverage new IT capabilities.
3. Determining reference architecture standards. Knowing what technology is needed to deliver the service levels required enables IT to document this infrastructure in the form of architecture standards. These standards should be expressed in terms of classes of technology needed to deliver each tier of services required. It should not document vendor products.
4. Identifying costs. If identifying the technology support requirements has tracked present and required investments to deliver each tier of service, enterprises are now prepared to look at the tradeoff of cost versus service delivery for each domain of service. This tradeoff is best facilitated through a spreadsheet tool, if possible, but the point of this success factor is for IT to have a business discussion with their business stakeholder about what they are willing to pay, literally or figuratively, for the services they require. Costs to deliver each service tier should be identified per tier.
5. Identifying applications to be replatformed. This should lead to decisions about moving applications from existing platforms to new, usually less expensive tiers. Enterprises always have a number of applications that are mission critical and must stay on an infrastructure that provides the highest level of service. Companies will realize the greatest business impact by focusing on applications receiving highly expensive, tier one-type services.
6. Assessing Financial Impacts. The collective value of this process becomes clear once a client adds the savings or cost avoidance realized application by application. For example, a large financial company facing data growth at an annual rate of 135 per cent, resulting in US$8 million in unplanned/unbudgeted costs targeted more than US$19 million in cost avoidance over a two year period. Within the first five months, it actually achieved US$3.3 million of this cost avoidance.
One of the residual benefits of developing a storage service catalogue and architecting to meet its requirements is that it helps enterprises prepare for the future. This approach not only helps companies lower costs and avoid unnecessarily early purchases but also helps IT predict upcoming infrastructure costs more accurately. And, this storage service catalogue approach helps customers prepare for new infrastructure capabilities, such as virtualization. Knowing the business applications service level requirements prepares IT to think in terms of delivering classes of service from a pool of IT capabilities.
The growing movement toward ILM is dictating that well planned and carefully designed tiered storage infrastructures be put in place to provide the underlying technology foundation for future storage infrastructures. Given the intense demands of the global marketplace, the time is now to start developing these advanced infrastructures.
Iain Anderson is Client Solutions Director of EMC Canada’s Technology Solutions Group. EMC Consulting, a division of EMC’s Technology Solutions Group, provides expert consulting services on the strategic, architectural, operational and managerial best practices for sustainable information infrastructure.