Issue 7, Dec 09/Jan 10


FOCUS IT REFRESH


puE as a function of annual utilisation profile applied to three different options topic


low utilisation


Current annual hours at each utilisation rate (IT load) Future annual hours at each utilisation rate (IT load)


PUE (IT load/total load for each utilisation rate) Option 1 - Reuse existing cooling system Option 2 - Adapt existing cooling system Option 3 - TCO Optimised cooling system


Figure 2: annual utilisation profile applied to Options 1-3


One of the distinct advantages of the IT refresh is that it may be possible to access historical data or trends of IT equipment utilisation


A standard industry metric for characterising the effectiveness and efficiency of a data center is the Power Usage Effectiveness (PUE) or its inverse, the Data Center infrastructure Efficiency (DCiE). Both are rations of Total Facility Power and IT Equipment Power and can serve as a primary or secondary baseline for options presented previously.


Figure 2 shows varying IT utilisation in terms of both percentage and hours per year, similar to a bin analysis for weather data. These utilisation rates are mapped against each of the options identified in Figure 1 to show representative PUE values.


As seen from this example data set, the Total Cost of Ownership (TCO) optimised cooling system will yield the greatest average PUE across the spectrum of utilisation hours. In this example, the cooling equipment selected, whether reused, adapted or new, will operate most efficiently when IT utilisation rates are increased in the post-refresh data center.


The utilisation data used to populate Figure 2 is most easily obtained through close co- operation with the IT professionals leading the refresh activities. Cross-functional information transfer can be an invaluable tool for analysis when TCO is the target metric.


Opportunities for addressing infrastructure changes during an IT refresh are based on a business case that is ultimately influenced by a fair and unbiased evaluation of options.


The concept of formulating options that are truly vendor-neutral and free from outside influences is no minor undertaking and requires a conscious, vigilant effort. It is key to develop options that do not artificially limit the spectrum of choices such that only one option is established as viable.


Conversely, it can be equally debilitating to generate a matrix of choices, where every possible permutation has been identified. The consequences of trying to validate too many choices can be ‘analysis paralysis’, which ultimately compromises the results.


One potential solution for establishing the correct due diligence choices is to conduct a ranging analysis. In practice, this would include the following typical options:


• OptiOn 1 MaxiMuM practical ExtrEME The development of Option 1 should be as uninhibited as possible and not limited to a given system or operational mode.


Extremes should cover the spectrum of possibilities and could potentially involve scenarios such as rerouting data center traffic for six months during a complete infrastructure retrofit.


Other potential extreme scenarios could include planning for a 100% increase in compaction at the board level, or pre- provisioning for a cooling technology that is not yet commercially viable.


Option 1 should generate the ‘what if’ scenario and act as the upper bracket of Options 2 and 3.


• OptiOns 2 and 3 OppOrtunitiEs BEtwEEn MaxiMuM and MiniMuM ExtrEMEs These options should represent the best mix of holistic tradeoffs that create the most desirable business case. Options 2 and 3 test the boundaries of an unbiased approach, and requiring more than one choice avoids the trap of considering only one possible solution.


5,000 hrs 2,000 hrs


PUE 2.00 PUE 1.80 PUE 1.30


2,000 hrs 4,000 hrs


PUE 1.77 PUE 1.70 PUE 1.20


annual utilisation profile High utilisation


0 to 20% 21 to 40% 41 to 60% 61 to 80% 81 to 100% Trivial Trivial


1,760 hrs 2,760hrs


PUE 1.70 PUE 1.62 PUE 1.12


Trivial Trivial


totals


8,760 hrs 8,760 hrs


PUE 1.80 PUE 1.70 PUE 1.20


• OptiOn 4 MiniMuM practical ExtrEME Option 4 defines the minimum level of ‘change in state’. This can include minimising the concentration, frequency or invasiveness of an IT refresh, establishing the lowest possible capital cost for an upgrade or potentially reducing TCO by outsourcing some services.


The goal of structuring the potential choices into a limited number of options is to help bound the problem without unduly limiting the flexibility necessary to see what is possible. Additional considerations that can be considered include:


• 10% savings on cost of initial build or 50% reduction in disruption during next major refresh;


• Lowest capital cost or lowest five-year TCO cost;


• 25% construction capital cost savings, or free up 25% of cooling capacity;


• 15% construction capital cost savings, or easy future capacity expansion of 25%;


• 10% construction capital cost savings, or provisions for liquid cooling in the future; and


• 10% construction capital cost savings, or using non-potable water instead of potable water.


The percentages, TCO metrics and baseline questions are only examples, but it is evident that holistic tradeoffs are required to assess the impact of any infrastructure upgrade.


The preceding discussion highlighted how a ‘clean slate’ approach to infrastructure upgrades during the IT refresh can help decouple existing limitations and constraints.


A holistic approach is necessary so that variables such as economics, schedule, energy, operations, manpower available technology and ultimately TCO can be fairly evaluated within the context of the refresh parameters.


Whenever possible, the IT refresh cycle should be embraced as a distinct opportunity, since it is one of the few times when a data center can most readily accommodate adaptations, expansions and fundamental changes. 


THE AUTHORS John Lanni, PE and Don Beaty, PE of DLB Associates www.datacenterdynamics.com 41


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60