Data centre topology choices affect mission-critical business system availability. IBM has looked at a methodology for ensuring business resilience by assessing the risks to the mission-critical business systems and then designing a data centre topology to mitigate these risks. This methodology is based on the IBM Business Resilience Framework, a framework that accounts for a wide range of concerns: from data centre facilities to business strategy and vision.

 Advances in information technology (IT) have led to its adoption by most business and government processes in our society. Applications serving individual consumers, such as banking or entertainment, have evolved to the point of complete dependence on IT systems. Many business applications are critical to the success of businesses deploying them, thus, continuous availability of business IT systems plays an important role in modern economy.

The data centre is now an essential part of IT infrastructure, containing servers, storage and operational management capabilities necessary to support mission-critical business systems. Whereas overall business resilience of a company is determined by many factors, the number and location of its data centres are some of the most important considerations in its mission-critical systems design. Given that data centre construction is expensive and facilities are expected to last for decades, poor selection of data centre locations can be both costly and disruptive.

Selection of data centre topology results from a process that starts with examination of the overall IT environment, including business and regulatory requirements the client has to deal with. An understanding of the risks associated with data centre facilities has to be developed. These requirements and risks must be considered in selecting appropriate data centre topology.

The impact of data centre topologies on business resilience is evaluated and the results are used to determine the number of active data centres and the topology needed, as well as the minimum separation between data centres necessary.

Our experience has shown us that the following methodology is effective in data centre topology design that addresses the client’s business-resilience goals:

1. Create a business risk profile that includes known threats and regulatory requirements, such as requirements for data centre separation.
2. Quantify the financial impact of those threats or regulations on the business.
3. Identify potential risks to strategy, technology, processes, organisation, applications and data, and facilities, and their relation to data centre topology risks.
4. Design data centre topology, capable of mitigating identified risks and ensuring business resilience.

Data centre placement plays a critical role in determining the company’s overall resilience. Factors such as geology, weather, security, power, water and access are directly related to risks associated with any particular site. Here are some of the factors most relevant to data centre topology:

• Power consumption: the growing use of distributed servers and server farms to replace mainframes has greatly increased electrical power and cooling needs of a data centre.
• Floor space: although distributed servers have become smaller, their numbers have increased greatly, requiring ever-increasing amounts of space.
• Server consolidation: in an attempt to address the first two items, server consolidation is seen as a way to reduce server space, power, and support requirements.
• Data centre consolidation: with more mergers and acquisitions, data centre consolidation helps reduce overhead costs by reducing the number of data centres.
• Network capacity: rapid expansion of IT-based services to customers requires ever-increasing network capacity.

Data centre facilities are subject to many types of risk – both natural and human-caused.

For the purpose of evaluating a location, the risks associated with it can be described as a risk profile. The geographic areas contiguous to the location that share the same risk profile can be described as part of a region.

All data centre facilities situated inside a region share the same risk profile and, for the purpose of data centre topology design, are said to be in-region (this is also applicable to regulatory requirements).

Those sites that do not share the same risk profile are considered out-of-region. We should point out that some risks cover larger areas than others— natural disasters such as Hurricane Katrina in 2005 or the North American ice storm in 1998 affected large areas. Out-of region data centre sites are generally located in different metropolitan areas, different states and even different countries.

All in-region risks should be mitigated through the use of a resilience strategy that fulfils all availability, recovery, business continuity and security requirements. To this end, the IBM Business Resilience Framework, illustrated in Figure 1 (see below) as a layered hierarchy, accounts for a wide range of concerns, with facilities as a key one:

• Strategy and vision: relates to strategies the business uses to complete day-to-day activities to achieve continuous operations. Examples include compliance, governance, availability, continuity and security strategies.
• Organisation: relates to organisational structure, communication channels, people skills and responsibilities. Examples include human resources, training, internal and external communications.
• Processes: relates to critical business processes necessary to run the business, as well as supporting IT processes. Examples include accounts receivable, accounts payable, incident and change management.
• Applications and data: examples include customer relationship management (CRM) applications, enterprise resource planning (ERP) applications and database and transaction-processing applications.
• Technology: relates to systems, network and industry-specific technology, necessary to enable business applications and data. Examples include host systems, workstations and IP networks.
• Facilities: relates to buildings, factories or offices, necessary to house production or service infrastructure and staff. Examples include data centres, office buildings and physical security operations.

In Part 2 of the series (coming Wednesday): IBM team identifies key measures in designing data centre topology for availability.

Authors: Richard Cocchiara, Distinguished Engineer and the Chief Technology Officer for Business Continuity and Resiliency Services at IBM
Dr. Hugh Davis, Lead Architect in IBM’s Global Business Resilience Consulting Practice
Doug Kinnaird, Executive IT Architect in IT Strategy and Architecture Practice at IBM

Related news: Equinix opens new data center in Los Angeles area
Related video: Connecting the dots in data center automation
Related white paper: Service automation: your path to a more efficient data center

Keywords: IBM, data centre, data centre topology, disaster recovery