Fundamentals of data centre life cycle management

SUCCESSFUL EXECUTION of a project depends in large part upon the process that pilots it through development and realization, from concept to commissioning. Common problems such as wasted time and money, or defects are often due to the lack of an overarching, shared process guiding all parties, clarifying responsibilities and communication.

A standardized process that meets the above general requirements will have the following characteristics:
Every activity necessary for completion of
the project is included.
Each step has clearly defined inputs and
outputs.
Every output is either the input to another
step, or is a final output of the project. No
effort is wasted on extraneous outputs that
do not contribute to the progress or
ultimate outcome of the project.
Every step of the process has clearly
assigned ownership responsibility.
There are no dead spaces between steps
and every step is linked. Once a step
has received all its inputs, it can complete
its tasks and make its outputs available
to other dependencies.
There are special “asynchronous”
functions that remain on standby
throughout, to systematically deal with
unplanned changes or defect correction.
A web-based tracking and status system
is accessible to all stakeholders, for
shared documentation, data, and reports.

Plan Phase
The plan phase should be the shortest and least expensive of all of the phases. With the potential for having the greatest impact on both the costs and capabilities of the data centre, it should determine the key project parameters of the physical system to be created, site selection, and the project process that will determine them. Organized and managed properly, projects will go more smoothly with fewer surprises occurring in later phases that may result in costly delays, large numbers of design iterations, and rework. Managed poorly, the phase can last for months and negatively impact the length of subsequent phases.

Design Phase
This involves the detailed design work required to translate the plan phase outputs into site-specific schematics and buildable construction documents. It is also during this phase that construction contractors are evaluated and selected. In a well-managed “design/ bid/ build” model, this phase can take about 8-10 weeks, but what specifically happens will vary depending on many factors including the type of construction method used, size/ scope of the project, as well as the particular needs, preferences, knowledge, and skills of the data centre project team.

Before the next phase begins, this construction team develops a project plan with a clear identification of critical path items, creates a commissioning plan, budget, billing plan, and a schedule. It is important for the construction team to work closely with the design team as design and construction documents are finalized to ensure these plans, schedules and budgets reflect reality. And the owner should be reviewing and approving this work as it is done.

Build Phase
The construction will commence the build phase in accordance with the project plan, and team members will focus on their assigned disciplines. Regular quality assurance and audits should be performed and reported. Weekly meetings should be held and reports generated as work progresses. The owner’s focus should be on the status of progress, quality performance, and identifying schedule risks early.

As systems are built, installed, and integrated together, a good opportunity is presented for training the facility operations team. It is the construction team’s responsibility to develop the punch list, as-built drawings, equipment manuals, sequence of operation documents, material safety data sheets, and warranty documents. This documentation combined with hands-on training by equipment vendors forms the source content for developing the emergency operating procedures (EOPs) and methods of procedure (MOPs). These documented procedures are a key aspect of an effective facility Operations & Maintenance (O&M) programme.
Operate Phase
This is the longest and costliest phase and often lasts 10, 15, or even 20 years. It is during this long period that the physical infrastructure is doing what it was intended to do: to house, power, cool, and secure IT servers, storage, and networking gear. The physical infrastructure must continuously function even as equipment ages, is serviced, and is eventually replaced.

Operations must be continuously maintained even as IT and business demands fluctuate. Some have likened it to maintaining an aeroplane while flying it. Failure is not an option, particularly since business success is so often tied to IT performance.

A properly designed, implemented, and supported O&M programme will minimize risk, reduce costs, and even provide a competitive advantage for the overall business. A poorly organized programme can quickly undermine the design intent of the facility putting its people, IT systems, and the business itself at risk of harm or disruption.

Assess Phase
Once the facility is in active operation, the Assess phase begins. Although concurrent with the Operate phase, Schneider Electric views the critical task of monitoring and formally evaluating performance as a separate and distinct phase of the overall life cycle - arguably the one most often neglected. Regularly assessing performance is an important element of an effective O&M programme. Understanding how operators and the physical infrastructure systems are performing on an on-going basis yields useful and actionable information including:
Degree to which the design intent and
facility objectives are being met by the
current infrastructure in operation
Energy efficiency of physical infrastructure
in supporting IT
General health and current risk profile of
the physical infrastructure
Current power, cooling, and space
capacities and use trends
Effectiveness and experience levels of
personnel
Facility operations & maintenance
program maturity and effectiveness
Assessing performance is most effectively done through the use of key performance indicators (KPIs), which are used to provide focus and drive program improvements. This yields several benefits, including the alignment of operational activities with business goals and providing positive reinforcement for innovation and process improvement.

Outsourcing
For each phase of the data centre life cycle, there are vendors who offer services to advise, manage, start-up, configure, train, assess, test, and operate. Procuring these services can offer several potential advantages including:
Benefiting from the experience and
expertise of others
Freeing up internal resources to focus on
other key tasks and responsibilities
Avoiding the natural learning curve and
the mistakes that might be expected when
tasks are being done for the first time
Providing an alternative means to achieve
goals when resources are lacking
internally to learn required skills or perform
some function.

It is important for owners and their management teams to be aware of and understand the availability of services for the facility in question; i.e., what is specifically offered, what service levels are possible, and what its costs are. Compare service providers and vendors in terms of capabilities, expertise, experience, coverage, and cost. Then evaluate all of this information in the context of what the facility internally offers today in terms of staffing coverage, skills, and experience. For those who lack the required expertise and resources, effective use of third party services can make owning and operating data centres easier and more efficient.