01:22 PM

SmartAdvice: Craft A Data-Center SLA So It Meets Business Needs

Outsourcing data-center work lets staff focus on higher-value activities, but pay attention to business metrics when drafting the agreement, The Advisory Council says. Also, RSS may be a useful tool depending on your's company's industry; and expect staff cultural challenges when switching from mainframe based apps to client/server or Web-based applications.

Editor's Note: Welcome to SmartAdvice, a weekly column by The Advisory Council (TAC), an advisory service firm. The feature answers three questions of core interest to you, ranging from leadership advice to enterprise strategies to how to deal with vendors. Submit questions directly to [email protected]

Question A: What service-level agreement provisions should we expect to be able to get on data center outsourcing?

Our advice: Data centers come in all shapes and sizes, and so do SLAs. With that in mind, the keys to establishing an SLA for a data center outsourcing project will include clearly defining the objectives and requirements; holding the service provider accountable for its work while also motivating it to stick to the agreement and meet deadlines; and monitoring the assessment metrics to ensure their fairness.

The data center exists for the sake of the business, not for the sake of technology. Therefore, the SLA must be based on the needs and expectations of the business user.

Data Center Outsourcing
Data center outsourcing enables internal resources to focus on high-value activities, and the organization on its core competencies. The SLA should ensure that this outcome isn't compromised. The most commonly encountered types of data center outsourcing include:

  • Platform-hosted services such as co-location and Web-site management.
  • Subscription services such as E-mail hosting and wireless messaging.
  • Storage services such as backup, off-site storage, and recovery.
  • Performance reporting, change control, and similar services.

The typical benefits touted by data center outsourcing service providers include:

  • Cost effective, incremental scalability.
  • Around-the-clock access using redundant infrastructure.
  • On-demand service delivery and load balancing.

The primary drivers behind data center outsourcing are:

  • Reducing costs.
  • Eliminating a function that's not an enterprise core competence.
  • Bringing best practices into the data-center operation.
  • Freeing internal resources to work on more critical projects.

Service-Level Agreements
Data center outsourcing can positively affect a company's top line (e.g., through improved service on customer-facing applications) and bottom line, and could help it achieve industry best practices. The SLA should ensure that the best-practices objective is reinforced; therefore, a number of criteria should be included in data-center outsourcing SLAs, specifically:

  • What specific services are to be provided?
  • How the provider will deliver these services.
  • Who will measure service delivery, and how?
  • What happens if the provider fails to deliver?
  • How the SLA itself can be changed over time.

Related Links
Sun Microsystems
Here are the steps to be followed in creating an SLA:
  • Identify service levels that your infrastructure needs, so that the SLA is comprehensive.
  • Draft the SLA so that it clearly defines the service provider's responsibilities.
  • Negotiate the SLA with the service provider, paying particular attention to what services are being guaranteed, how they will be measured, the process for realizing agreed-upon remedies, and the amount of time the service provider has to correct problems.
  • Implement SLA measurement and enforcement tools and processes, to ensure that every SLA criterion can be measured and enforced as soon as the service is delivered.
  • Enforce SLA compliance, and identify and resolve problems that arise.

General SLA Guidelines
Some useful guidelines to keep in mind when creating SLAs include:

  • Ask only for what you need, and identify and protect the most important and critical organizational assets.
  • Define your criteria, and how you will continuously monitor them.
  • Cover best- and worst-case situations in the agreement, and make sure any penalties are fair.
  • Choose rewards with care, and keep them to a minimum.
  • Demand continuous improvement from the provider.
  • Get both staff and management buy-in.
  • Assign SLA enforcement responsibility.
  • Have an exit strategy should the need arise.

With proper planning, a company can create a data center outsourcing SLA that will meet its organizational needs. Be clear about best- and worst-case scenarios and draft the SLA in partnership with the provider, emphasizing the processes that ensure that terms are being met and that service continually improves.

-Sanjay Anand

Question B: What is RSS, and what, if anything, should we be doing with it?

Our advice: What industry you're in may dictate how aggressively a company pursues RSS technology. Content aggregators and publishers are already using it heavily, but there are some open issues to be weighed.

Consider the purpose and reasons for your organization using RSS, and avoid jumping on any bandwagons. Educate your staff and your users about RSS' benefits and value. Evaluate opportunities--possibly start a blog--and understand the requirements for supporting it. Finally, plan accordingly.

RSS Defined
RSS (variously defined as Rich Site Summary, RDF Site Summary, or Really Simple Syndication) is a method for delivering news or Web content from a syndicated source to subscribers, a multipurpose extensible metadata description and syndication format.

Although RSS has been around since 1999, multiple versions still exist, and some formats are unreadable by some tools. In addition, different groups supporting different versions each claim theirs to be the standard. Such confusion hasn't built either legitimacy or confidence amongst users. RSS readers (e.g., FeedDemon, Pluck, and NewsGator) are, for the most part, a separate component from the browser. We think that RSS readers should be part of the browser, such as in Mozilla's Firefox and Opera Software's browser.

Related Links
Opera Software

Mozilla Firefox

OASIS RSS Technology Report
Bloggers have long known the benefits and value of RSS, since those sites are routinely updated with up-to-the-minute posts. Portals such as Yahoo, and major news sites (e.g., New York Times and Wall Street Journal) have had the ability to incorporate RSS feeds for some time, because it enables news stories and headlines to be updated automatically in near real-time.

Ever wonder what those orange-colored logos with the letters XML are on your favorite Web sites? They're links to RSS feeds: Subscribers automatically receive site updates because the publisher registered the site with an RSS aggregator and tagged RSS documents with XML codes that describe the type of document and where it resides.

RSS Is Not Push Technology
On the surface, RSS appears to resemble the push technology of vendors such as PointCast, but there are differences. Push technology was built on the broadcasting model, where providers fed content to desktops, accompanied by lots of ads; RSS places the choice of content in the hands of the user, aggregating the choice of readers.

For content providers and publishers, RSS provides an easy way to syndicate data. For browser vendors, it's a means for regaining control of the online experience. For consumers, it lets them better focus on content without being distracted by pop-ups and spam. This has some nicknaming RSS "the TiVo of the Web."

-John C. Sinclair

Question C: How should we deal with the cultural and skill-set changes needed when moving from mainframe-based applications to client/server and Web-based applications?

Our advice: Expect cultural challenges among the management and technical advisory staff, the technical support and operations staff, and among the software developers as they become acclimated to new roles and concepts.

Management Tips Be sensitive to the possibility of five pre-existing mindsets as they might affect the selection process:

Moving from MVS to anything else is called "downsizing." Both Unix and Windows have demonstrated the ability to handle huge and equivalent transactional loads.

Moving from MVS means compromising availability. In fact, non-MVS systems can be configured for absolute availability, by which we mean around-the-clock transaction processing for multiple consecutive years, capable of surviving fire, flood, or terrorist attack without loss of service and without data corruption. The team should be open to a collection of different technical approaches used to achieve this result, such as clustering, storage networks, and rolling upgrades.

Moving from MVS means compromising security. Unix and Windows systems have often been configured to make access as easy as possible, and therefore vulnerable to hackers. Properly managed versions of Unix and Windows can be kept as secure as MVS.

Unix is more virtuous, because it's open. Windows is just as open, if you define open as permitting easy porting of applications across multiple hardware vendors.

Linux is free. Cost must be computed for the whole environment, including middleware, security, and system-management applications, which make the cost of Linux comparable with other alternatives.

Be aware of potential concerns staff members may have about losing their status because Window is easier to use and maintain than MVS. There are still specialized technical skills required in a Unix or Windows data-center environment, so while many users know how to install Windows, not many people know how to configure complexes of Unix or Windows systems for absolute availability, nor how to design tight security or maintain portal servers. We recommend you tell the technical staff early on that they will not lose their status.

Software Developers' Concerns They will face several issues, such as a) approaches to the enhancement of the new applications, b) development speed, c) the extent to which users are empowered to program independently, and d) perception of the human interface. With those in mind, consider these:

  • If your current MVS applications are more than five years old, there's a chance you've been making modifications directly into the vendor's code. You don't want to do that moving forward. Modern applications are written with clean interfaces that permit adding enhancements by surrounding the vendor's code (seen as a black box) with a periphery of objects of your own. This permits the easy installation of new releases of the vendor's application.
  • If your developers have been making modifications directly into aging code, they may be deliberate in their approach, yet the new tools allow cleaner objects around the periphery and will speed the pace of development. You may want to train a few of them in Rapid Iterative Prototyping techniques, and let the RIP team be a model for new request-to-production response times.
  • If your current applications are more than eight years old, a simple request for a new report becomes a developer issue. With the easy availability of multidimensional data stores and user-friendly report writers, developers should train users to do it themselves. You shouldn't encounter any job-protection reluctance. It should be noted that the potential for online analytical processing (OLAP) may bear on your technology decision, since OLAP is far less expensive in a Microsoft SQL Server environment.
  • If your company is ready to emerge from a human-interface system to one where each person has a personally customized cross-application interface, then this shift should be factored into the new application selection. In addition, the development staff should avoid retrofitting the portal technology at a later date.
  • -Wes Melling

    Sanjay Anand,, TAC Expert, has more than 20 years of IT and business-process-management experience as a strategic adviser, certified consultant, speaker, and published author. More than 100 personal clients, large and small, have included companies from a diverse array of industries and geographies, from academia to technology and from Asia to the Americas. Often referred to as a "consultant's consultant" for training and mentoring skills. He is author of books "The Sarbanes-Oxley Guide for Finance and Information Technology Professionals" and "J.D. Edwards OneWorld: A Beginner's Guide."

    John Sinclair, TAC Expert, has more than 26 years of experience spanning a variety of industries. His last 11 years have focused on process, project, and information management in consumer-goods and clinical-trials environments with experiences in document and content management, systems validation, FDA compliance issues, and 21CFR-11 legislation. His strengths include research and analysis and evaluating the business use of emerging technologies. He's certified as a Project Management Professional and is a graduate of the Society for Information Management's regional leadership program.

    Wes Melling, TAC Expert, has more than 40 years of IT experience with a focus on enterprise IT strategies. He's founder and principal of Value Chain Advisors, a consulting boutique specializing in manufacturing supply-chain optimization. He's been a corporate CIO, a Gartner analyst, and a product strategist at increasingly senior levels.

    We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.
    Comment  | 
    Print  | 
    More Insights
    Copyright © 2020 UBM Electronics, A UBM company, All rights reserved. Privacy Policy | Terms of Service