Advertisment

What is the right management approach for DCs

author-image
DQI Bureau
New Update

Data Centres in India

Advertisment

Data centres in India have been growing at a fast space since the last 10 years or so. Ever since the National E-Governance (NeGP) project was kicked off in May 2006, it has created a platform for multiple data centres across Indian states to support its state-wide e-governance projects. An initial investment of `1600+ crore is planned over the next 4-5 years. This is apart from many private players who have invested in some of the largest data centers. Organizations are making a big shift from a distributed IT setup to a more manageable, efficient centralized model, leading to consolidation of branch and remote IT resources into fewer, but larger data centers.

Though we are in the early stages of cloud adoption, cloud service providers will also be a key contributor to the growing need for more data centers. As per Gartner's May 2012 forecasts over the next 5 years, the growth of servers would be at a CAGR of around 6% while for storage systems is about 18%.

TechNavio's analysts have forecasted that the data center equipment market in India may grow at a CAGR of 10.4% over the period 2011-2015. As per Indian data center market review 2012, data centre market in India is expected to grow at a CAGR of 22% and will touch `6,500 Crore by 2016. Looking at the Data center focused projects in India, this may be possible.

Advertisment

Government of India's UIDAI project is already building data centers of 50,000 sq ft, each in Manesar and Bengaluru as a part of its Central Information Data Repository (CIDR) Infrastructure. In fact, global companies have envisaged this segment in India as a growth market and have been investing in Indian data centers over the last year or so.

Data Center Life Cycle (DCLC)

The immediate view for developing a data center is usually focussed on primary aspects such as space and size requirements while the long term view on sustenance and energy management often comes after few months of operations. DCLC approach is expected to help users consider the necessary steps in taking a holistic view on data centers, which would eventually help them from a design, build, operations and continued sustenance perspective.

Stage1-Business Requirement Analysis

Advertisment

Business Requirement Analysis is the first and the most critical stage which will decide the overall space and sizing for the Data center. In this stage, we derive the following:-

  • Total compute requirements phased over a period of next 3-5 years.
  • Centralized or distributed requirements at various locations.
  • Primary, DR, and BCP site requirements.
  • Ascertaining the capability of existing hardware available to support existing/new requirements.
  • Estimation on space/power requirements based on compute requirements and technology refresh roadmaps from OEM's.
  • Design for a green field data center or a retrofit DC where existing one is already available.
  • Finally, documented plan approved by the Senior Management with Project completion timelines.

The output of this stage is a complete analysis of the business requirements, computational needs, business projections, number of devices needed and projections for the next few years.

Advertisment

 

Stage2-Planning and Design

Planning and design is the strategic stage in which the following is usually considered:

Advertisment
  • Planning for locations-Decision on where in the country to host depending upon economic & political situation, environmental factors, weather, seismic conditions, power availability, etc.
  • Data Center Tier (TIA 942) and uptime guideline considerations are generally factored to showcase efficient designs. This has to be carefully balanced with the capital requirement and real uptime needs.
  • In-house designed or need of a design agency.
  • Determining the size of white space and other areas depending upon compute requirements including growth and future expansion, non-DC, support & utility area space, etc.
  • Power Supply Options-Power utilities available, primary feed, backup feed possibilities, DG set size, environmental factors, use of green energy, etc.
  • Telecom connectivity including diverse routes and service providers.
  • Statutory compliances to be factored. (for eg, banks need to comply to RBI norms, etc)
  • Budgets and financial approvals followed by the developing request for proposal bids, evaluation process including the service level compliance depending upon needs.

Stage-Data Center Commissioning

Data center commissioning stage would see the actual delivery of the project. This stage will cover the following aspects:-

Advertisment
  • Vendor Team mobilization will include how early the work starts. Usually, this stage is found to delay the project for want of key personnel required to kick-start the work.
  • Equipment delivery & installation practice will help in how fast the work gets completed. It is necessary that each OEM gives an undertaking to the full cooperation for installation and support, usually over a period of 5-10 years for supply of equipment & its serviceability during its usable life.
  • One has to carefully monitor the work to avoid Time-overrun and ensure that the project is within the control, both of the executing vendor and the owner of the Data center.
Advertisment
  • Project would be completed on time if both the Vendor team and the Project owners collaborate and work for each other instead of the usual blame-games which ought to delay projects, sometime also leading to cost-overruns by the implementers.
  • Finally, a well documented project covering the warranty, support and handover documentation would enhance the credibility of the implementers and help the customer, i.e. the Data center owner in achieving its objectives.Stage-Data Center Operate, Sustain, and Enhance Stage

Stage-Data Center Operate, Sustain, and Enhance Stage

The DC operate, sustain, and enhance stage truly reflects on the best practices followed by the data center team on its ongoing healthiness and Continual improvements. This stage will factor the following:-

  • Ensuring that the Operations team is well qualified to run the Operations on a day to day basis, 24x7 management of equipment & people shifts without uptime degradation/ downtime, ensuring availability and manageability using right skilled team.
  • Load and thermal management of the data center avoiding hot spots and excess cooling.
  • Continual OEM support.
  • Management overview using standard operating procedures, checklists, SLA, and task compliances.
  • Reviews-Internal teams, OEM, other support vendors.
  • MIS reports for customer/senior management confidence in services delivered.
  • Actively considering cases of improvements and its implementation, preferably on a pilot basis.

 

 

Stage-Re-assessment for Changes and New Requirements

Re-assessment stage should be essentially planned right from the first year itself rather than waiting for couple of years. It should factor the changes that a data center needs in terms of new requirements, changes to the environment like Cooling, Power requirements, air-flow, ageing of equipment, changes in efficiency such as increased power consumption or inability to deliver the work as per the specifications. As a part of the standard operating procedures, this stage will help improve the utilization and keep checks on the equipment. Based on the behaviour of equipment, it can be possible to predict failures due to overheat, high current consumption, excess fuel consumption, sudden variances, etc. This stage is essentially a regular audit of operations and requirements to ensure that users don't enter a stage of surprise either at the requirements, equipment, availability, space crunch, obsolescence, etc, and a fair prediction can be made, an essential element for Senior Management intervention and action on timely basis.

Data Center Maturity Model (DCMM) Tool for Planning & Improving Energy Efficiency across the Entire Lifecycle

Data center owners/operators interested in improving their data center's energy efficiency and sustainability will gain from the DCMM tool developed by the Green Grid in 2010. This model provides clear goals & direction for improving energy efficiency and sustainability across all aspects of the data center.

DCMM touches upon power, compute, storage, cooling, networks, other facility, other IT. Users can benchmark their current performance using the data center maturity model equaliser. By determining levels of maturity and identifying the steps/innovations required as part of the DC & IT strategy, achievement towards greater energy efficiency/sustainability improvements are possible.

Adopting DCMM can help companies use data center assets more effectively, reducing their capital & operational expenditure, as well as minimizing their environmental impact and improving their corporate/social responsibility standings. Data center owners/operators and senior managers can use the DCMM tool for direction, self-assessment, and informed decision making. They can determine where they stand in relation to current best practices and understand the roadmap of the industry.

For DCMM, the following steps are defined in moving from a Low efficiency to the highest achievable efficiency through progressive use of best practices:-

  • Level ‘0' (Has low or no progress)
  • Level ‘1' (Started to adopt part best practices)
  • Level ‘2' (Has adopted best practices)
  • Level ‘3' & ‘4' (It is better than most others)
  • Level ‘5' (Visionary, is likely to achieve targets over a 5-6 year period)

So level 0 data center is a one which has taken no efficiency improvement measures while level 5 is the ultimate in energy efficiency. Levels 3 through 5 represent future capabilities toward which the industry should collectively move and innovate.

DCMM tool simply provides a yardstick with which to measure progress, unlike other ‘Codes of conduct' or rule sets which may sometimes end up as counter-productive.

Conclusion

While one would like to consider a proprietary approach to developing and managing data centers, it is best suited for the organisation to select a proven method. The description of a data center lifecycle is a process which covers various aspects from strategy, to design, implementation and continual improvement which has been adopted from the more conventional Plan-Do- Check- Act (P-D-C-A) aspect of the quality standards and is universal to either a service provider data center or acaptive data center.

 

 

Advertisment