Introduction
Cisco provides help to improve a business operations energy efficiency for a range of business processes and activities. Whether a user has a green agenda or needs more power capacity to scale the business, Cisco can help. Cisco primarily supports a customer defined green agenda through improvement in IT operating efficiency. Other areas of focus include better facilities operating efficiency, support for e-waste recycling programs, and reduction in peripheral materials such as cabling.
This document in intended to help customers and partners clearly understand what, where, and how Cisco can provide assistance. This document is not a scope of work (SOW), but it does provide insight into some of the capabilities that can be included in one. Sample SOWs are available from your Cisco service representative
Overview
There are many ways to group the solutions Cisco can offer to support energy efficiency as a planning, design, and operations consideration. This document employs a solutions framework that covers the following areas:
• Products
• Architecture
• Services
• Operations
• Partnerships
To navigate these solution areas, it is important to have at least a baseline understanding of efficiency analysis, electrical generation, emissions resulting from electricity, and IT operations. Cisco provides learning resources for these areas and more at http://www.cisco.com/go/efficiency.
All the solutions sections have a similar framework, each covering the attributes of a given solution. These attributes are grouped accordingly to the following structure:
• What is the solution definition?
• Where is the solution applied?
• Why is the solution compelling?
• How is the solution deployed?
Further information about these solution areas is available from Cisco account and service management teams. Cisco also hosts a planning and reference site to support the Cisco® Efficiency Assurance Program, at http://www.cisco.com/go/efficiency.
About the Cisco Efficiency Assurance Program
The Cisco Efficiency Assurance Program is a reference, learning, and planning program delivered through an interactive Web 2.0 interface hosted on Cisco.com. This portal, found at http://www.cisco.com/go/efficiency, showcases ways that Cisco can support a customer's green agenda by improving energy efficiency.
This program provides a path for Cisco customers to estimate the effects of IT operations financially and with sustainability in mind. Within the program is the Green Data Center Model planning tool, which calculates the monetary benefits of network-based virtualization and the carbon dioxide resulting from IT operations. Several other planning tools can be found at this site that allow users to build the use case for more sustainable IT operations including the Product Efficiency Calculator.
Although there is still much confusion today over green metrics, Cisco is helping ensure that all entities are working toward the most relevant metrics for energy efficiency related to business operations. While this program is not a comparative metrics tool, the metrics employed adhere to The Green Grid metrics of power usage effectiveness (PUE) and data center infrastructure effectiveness (DCiE).
Document Assumptions
• Green is primarily a social and political term and will be referred to in this document as such.
• Energy efficiency refers to basic units of energy and equivalents thereof.
• Sustainability as a scientific term will be referred to in this document as a point of consideration in business decisions that reflects a more accurate cost of using finite natural resources.
• Ecologic considerations in this document will be balanced with economic considerations to help ensure that solutions adoption is viable.
• Efficiency of energy use related to human activity will be the primary focus on this document, with recycling and reuse considerations as secondary priorities.
Efficiency Analysis
There are many forms of energy and even more methods for measuring the efficiency of energy use to satisfy a given workload or workload grouping. Forms of energy, including , but not limited to, kinetic, potential, thermal, gravitational, sound, light, elastic, electromagnetic, chemical, nuclear, and mass energy, have been defined to explain all known natural phenomena.
The discussion in this document is confined to the analysis of useful work derived from electric, thermal, and mechanical energy. Said another way, electrical use, heat removal, static air pressure, and ultimately the carbon dioxide (C02) equivalents resulting from IT and supporting facility operations will be examined.
In this document, the concept of useful work does not derive from entropy. Rather, it is a subjective metric that must ultimately be defined by a user. For example, an IT system that supports an application such as email can be directly compared to another IT system provided that the application architecture and feature sets therein are identical. The moment different feature sets are added or removed from an application support model, the user must determine what is most useful to his or her particular use case.
In the case of data center efficiency analysis, no industry-adopted, globally recognized standards exist that comprehensively define efficiency. However, there are emerging metrics that will be cited in this document. Any emerging metrics, formulas, and analysis will take into account energy efficiency, defined as follows:
Efficient energy use, is wasting less energy to provide the same level of energy service. Efficient energy use is achieved primarily by means of a more efficient technology or processes. As with any new technology or process, changes in individual behavior can be altered to also waste less energy.
Cisco Energy-Efficient Solutions
Cisco provides solutions that, when implemented to their full potential, can provide significant energy efficiency gains across a range of human activities. For the sake of this document, Cisco is primarily focused on the implementation of network-based solutions that improve a business's operating efficiency. In any of the solutions Cisco presents, qualified metrics will be employed to provide credible benefit analysis. Where industry metrics do not exist, Cisco will provide best-in-class or emerging industry metrics from standards bodies, industry consortiums, governmental bodies, and other third-party bodies. Cisco will not employ metric frameworks resulting from vendor-financed analysis bodies.
An estimate of the electrical efficiency improvements across a data center are provided as an example of some of the solutions Cisco and partners can provide. (Table 1).
Table 1. Cumulative Estimated Savings from Deploying Cisco Energy-Efficient Data Center Solutions
* Figures are normalized estimates based on Cisco customer operative profiles between 2006 and 2008; the percentage is a total blended reduction across data center infrastructure (IT and facilities).
Global Factors Promoting Energy Efficiency
Cisco has several teams focusing on external energy-efficient solutions development, including engineering, professional services, technical solutions marketing, and customer advocacy teams. Cisco is also taking steps to aggressively reduce its impact as a user of resources. Cisco announced in June 2008 that it will seek to reduce its carbon footprint by 25 percent from 2007 levels by 2012. This percentage is an absolute reduction across employee travel and owned and leased real estate. Cisco's strategy for reduction involves using its own technology to meet this public commitment. Given that Cisco's footprint is almost 80 percent lab and data center, the solutions developed to support an in-house reduction will have relevance to a customer's IT management in support of a green agenda.
Figure 1. World Energy Consumption by Country, 2006
Source: Energy Information Administration, U.S. Department of Energy, 2006
The Economics behind Energy Efficiency
For each of the solutions, Cisco provides a business case as the first of several data series to consider in adopting more efficient operating models. Given the volatility of global energy and the steady increase in cost for all forms of commercially available energy, the business case for energy efficiency has never been stronger.
Gartner projected in 2007 that by 2012, energy will be the second or third highest cost for most large enterprises globally.
Figure 2. Historical Cost of Coal, the Primary Electrical Fuel Source Worldwide
Source: International Energy Agency (2006)
The business case analysis in this document primarily relies on planning tools that have been developed to support the Cisco Efficiency Assurance Program. These metrics include:
• Cost per kW/hr
• Cost deferment of new infrastructure, including server, storage, and network
• IT asset utilization
• Electrical systems utilization
• Operating efficiency defined with customer input
The business case for each use case varies, but these metrics provide a sound analytic foundation by which different approaches to IT services can be measured. In most cases, a business case for energy efficiency involves a before-and-after comparison. Although there are no industry-accepted standards bodies today that provide metrics on operating efficiency for IT systems, Cisco seeks to provide the most practical metrics available today.
Ecological Considerations for IT Operations
Cisco does not purport to be a recognized environmental advocate. However, Cisco has taken steps to build on its strong history of corporate stewardship and responsibility where global climate change is concerned. Cisco currently has investments in resources across its executive, engineering, customer advocacy, technical solutions, and brand marketing teams. All green subject-matter experts, including the authors of this document, ultimately report to the Cisco EcoBoard operating committee to help ensure sound governance, scientific accuracy, and credible analysis to support a customer's green agenda.
Some users may not be aware that the IT industry accounts for a sizeable environmental impact in addition to being an enabler for more efficient operations. Gartner provided analysis in a report released in 2007 that seeks to quantify the greenhouse gas (GHG) emissions impact of the IT industry. Gartner estimates that IT accounts for 2 percent of all anthropogenic GHG load in the environment (Figure 3).
Figure 3. Recent Analysis Suggests That IT Industries Contribute 2% of Global GHG Load, the Same Percentage as the Airline Industry
Source: Gartner Symposium/ITxpo 2007
Many IT professionals may be surprised to learn that the contribution of the IT industry to the problem of climate change is similar to that of the airline industry. This result occurs mainly because the most common fuel source for electricity worldwide is coal, and coal is one of the most inefficient fuel sources available today. However, coal is also the least expensive fuel source under current accounting, and there are ample reserves worldwide. Estimates of coal efficiency from extraction to delivery suggest that only 7 percent of the embedded energy of coal is delivered to a wall outlet.
Figure 4. Thermal Energy Includes All Fossil- and Biomass-Based Fuel Sources
Source: Energy Information Administration, U.S. Department of Energy, 2006
Figure 5. Per-Capita Electrical Use Worldwide
Source: Energy Information Administration, U.S. Department of Energy, 2006
Energy Use in the Information Technology Industry
To prioritize the use of energy related to IT systems and services, it is necessary to understand where energy is used within the larger context. IT energy use is primarily related to fixed assets such as real estate. To some extent, IT has a role in mobile energy use: for example, technologies that provide alternatives to travel, such as virtual presence, web collaboration and the measurement of mobile energy use.
Figure 6. General Breakdown of Energy Use Worldwide Showing a Half Going to Fixed Assets Such as Buildings
Source: BOMA, 2005
Although there are disparate claims as to what power goes where, most estimates today point to a roughly even split between centralized data centers and distributed computing desktops (Figure 7). Credible research on the actual breakdown of energy required within the entire IT industry is difficult to find as definitions of applied technology are in constant flux. For example, a data center today in many ways has a very different definition than a data center built in the 1980s. They are both central repositories for data, but the systems they consist of have changed dramatically. Most users choose to start with the data center because of its centralization of energy issues. However, to improve operating efficiency of energy use related to IT, all sources, desktops to data centers, should be considered.
Figure 7. Energy Use Specific to Computing Worldwide Showing Half Distributed and Half Centralized
Source: Cisco market Research Analysis, 2006-2008
Each of these computing environments is surrounded, supported, and enabled by a range of infrastructure, including networking, mobile devices, monitors, storage, and purpose-built facilities. According to Cisco market research, a general breakdown of the IT portion of data centers as of 2008 shows servers as the primary energy users, with storage catching up quickly; network systems, growing in their energy requirements, are a distant third (Figure 8).
Figure 8. Example of Electrical Allocation for IT Systems in a Typical Enterprise-Class Data Center
Source: Cisco market Research Analysis, 2006-2008
Components of Operating Efficiency
The term "component" can refer to the parts that make up a product if your interest is device efficiency, or it can refer to a product within a larger IT architecture if IT efficiency-that is, system wide efficiency gains across an enterprise-is your goal. Both types of efficiency are important in improving operating efficiency, but architectural changes have the most significant effects. Although this document does address device efficiency and encourages customers to specify highly efficient products, the main focus of Cisco Energy Efficient Solutions is on efficiency gains in IT and supporting facilities architectures.
Efficiency Metrics
There are many emerging metrics to assess green capabilities. Cisco has traditionally employed a number of metrics to help users determine the best product for the job they want it to perform. Cisco follows the best practice of including the work unit to support use cases in assessing electrical, thermal and carbon metrics. Cisco does not engage in product comparisons with competitors unless an industry-adopted standard metric is defined or cited with which Cisco could make a comparison.
Cisco is working toward solutions to simplify this complex process through its participation in The Green Grid. This consortium has already delivered a first level of metrics to define efficiency in the data center: PUE and its inverse, DCiE. Cisco considers product comparisons, which may only bring incremental point improvements, of lesser value than working with customers and partners to provide real solutions to global problems. To that end, Cisco is focused on the greatest operating efficiency improvements, which occur at the architectural level.
Consider the operating cost of a 1000-kW computer room air-conditioning (CRAC) system with 60 percent efficiency at US$0.10 (U.S. average for power) per kWh for 1 year at a 50 percent load level:
Operating cost = 1000 kW x (1/0.60 efficiency at load level) x 8760 hours x US$0.10 x Load level 50%
The Green Grid metrics for energy efficiency in data centers.
Product and System Comparison
Applying a power ratio in a product comparison for a use case is complex, but can be done. In the case of networking, a simple power-per-port metric for single-function equipment, fixed switching, and line card modules within modular switching can be applied. The power-per-port metric would then need to be assigned to a use case, except for products with multiple feature sets, which would need multiple use cases. A framework analytical approach would include:
• Watts per port
• Use cases to be defined by the users
• Broad architecture, including the computing and storage resources defined
• Use cases' allocated power supply requirements across the architecture
• Power supply requirements compared at the power supply and chip levels
• Load variance at multiple operating stages to be normalized
• Blended work unit defined by the use cases supported
• Blended work unit contrasted with the power supply requirements
To date, there has been no industry forum or clear immediate customer need to compare what are typically minor variances at the hardware level. Most modern IT equipment uses highly efficient switched-mode power supplies (SMPSs) when it is loaded properly. The greatest value in increasing operating efficiency lies in the IT architecture because it is developed to support applications. Redesigning applications should take into account its impacts on the infrastructure. The facilities that are designed to support IT infrastructure also need to be reevaluated to achieve higher efficiencies.
Cisco provides electrical efficiency curves for the power supplies that support many Cisco products through the Efficiency Assurance Program Product Efficiency Calculator at http://www.cisco.com/go/efficiency (Figure 9)
Figure 9. Cisco Efficiency Assurance Program Product Efficiency Calculator
To improve operating efficiency as a user of resources, Cisco IT is implementing the Cisco Efficiency Assurance Program. Specifically, Cisco Network Data Center Services (NDCS) teams are working through an initiative to benchmark and improve upon operating efficiencies across 52 data centers worldwide. While many of these data centers more closely resemble large labs, 8 are considered production data centers. Cisco anticipates a 17 to 20 percent normalized reduction in total power allocation to labs and data centers between 2008 and 2012 through electrical efficiency gains (Figure 10).
Figure 10. Example of an IT Asset Utilization Template Under Consideration by Cisco IT to Improve Electrical Operating Efficiency for Data Centers
Data centers and labs account for approximately 80 percent of Cisco's total energy use worldwide. Cisco's commitment to an absolute reduction across employee travel and owned and leased real estate will be accomplished primarily by using Cisco technology with Cisco IT having the largest role to play.
Energy Monitoring and Management
The Cisco EnergyWise solution is a new energy management framework that will enable IT to measure and fine-tune power usage to achieve significant cost savings. The Cisco EnergyWise solution focuses on devices connected to a Cisco network ranging from power-over-Ethernet (PoE) devices such as IP phones and wireless access points to IP-enabled building and lighting controllers. It uses an intelligent network-based approach that enables IT and building facilities departments to understand, optimize, and control power.
• What is the solution?
– A Cisco network-based energy management solution
• Where is the solution applied?
– Early stages focus on IT infrastructure; later phases focus on building infrastructure
• Why is the solution compelling?
– It provides an open standard to monitor and manage energy across any component that can hold an agent or accept a command-line interface (CLI) shutdown command, and the application is free
• How is the solution deployed?
– Inherent to Cisco IOS® Software
EnergyWise is a power management framework designed to measure power consumption and optimize power use, resulting in effective delivery of power across the enterprise. IT professionals can quickly optimize the power consumed in a building, resulting in immediate cost saving and a clear return on investment (ROI).
EnergyWise measures current power consumption and can automate and take actions to optimize power levels and report how much power is being consumed to demonstrate cost savings. After power consumption is understood, Cisco EnergyWise network protocols can be used to regulate power use. Energy consumed per location can easily be determined with a realistic view of power consumed per wiring closet, building floor, or campus building (Figure 11).
Figure 11. EnergyWise Optimized Power Delivery and Verification
The EnergyWise network is used to intelligently and proactively manage power consumption and consistently enforce policies to reduce energy consumption. EnergyWise provides the capability to monitor, manage, and reduce energy use by providing visibility into electricity consumption and the capability to switch devices from always on to always available based on business needs. EnergyWise offers orchestration and coordinated power management using the Cisco network for scalability and communication. For example, when an employee enters a building, a series of events can take place, enhancing building operating efficiency. The employee's badge access can trigger power-up of the office phone, wireless access point coverage, computer bootup, and setting of the office temperature to an appropriate value. Obviously, the EnergyWise user saves energy by powering off components when they are not needed.
In many cases, individual management systems are dedicated to different types of device in a building: management systems for building controls, another for phones, another for access points, and so on. Today a large number of systems need to be integrated to orchestrate events for power management. Integration of disparate systems can be difficult to achieve and not always implemented. EnergyWise network-wide policies can control device power management, eliminating the need for myriad systems to integrate and coordinate with each other. Orchestration is a primary benefit of this scenario, and it is the Cisco network acting as a proxy of information that allows systems to communicate in a synchronized fashion, reducing complexity and costs and enabling power savings. Figure 12 depicts a typical EnergyWise Cisco network, including the management layer and EnergyWise enabled endpoints.
Figure 12. EnergyWise Enabled Network
The Future of IP-Based Energy Management
EnergyWise works conjunction with other Cisco products, and partner products will ultimately extend into a common framework that includes control and optimization of all energy-using equipment within buildings. These converged systems include heating and air conditioning, lighting, and security, among others (Figure 13).
Figure 13. EnergyWise Intelligent Network Framework
EnergyWise enables intelligent load management; it makes intelligent dynamic load shedding possible. With EnergyWise, customers can monitor and control power during periods of electric grid instability and peak power events. EnergyWise is the first step toward making use of smart loads in network equipment. This step is valuable because buildings consume about one-third of the energy used in the United States. Electricity accounts for almost 80 percent of the energy costs in buildings.
Data Center Network Architectures
As stated in the previous section, data center efficiency must be quantified at the systems level and aggregated into a site-wide measurement. The position of the network is unique among the IT infrastructure segments. The network is the ubiquitous platform in the data center and is ideally suited to bridging the gap between facilities and IT.
When considering network architecture, a planner should not confine the scope to switches, routers, load balancers, and firewalls. Where possible, planners should extend the touch of the network to include as many relevant data sources as possible to monitor power efficiency, such as power, temperature, and humidity monitors; uninterruptible power supplies (UPSs); CRAC and computer room air handler (CRAH); branch circuits; power distribution units (PDUs); and specialized sensors that are typically found inside or close to the main floor of the data center. Including these components is non-disruptive and cost effective in terms of connectivity because they can be included as part of the existing or normally planned cabling structure.
In addition to the efficiencies gained by using the network as the physical middleware between IT and facilities, significant opportunities for efficiencies exist in network design to support various services. If application delivery, security, access and identity, storage, and computing are considered services, the network can have a profound effect on the cumulative efficiency of an operation. The following sections explore the implications of the right network architecture on power, cooling, and management. Migration strategies and best practices are also suggested.
The Network Effect
The network effect refers to the consolidation of services or portions of services provided by standalone components into a ubiquitous network-based platform. The dominance of this type of consolidation has been clear since the inception of switching and routing. Print, storage, and file services are obvious examples of services consolidated on the network, and this trend now encompasses enterprise-wide application delivery, security, web, identity, and access services. When analyzing the ratio of power used to service delivered, an important first step is to consider the effects of the network across a broad range of disparate infrastructure and service levels. To illustrate the network effect, the primary focus here is on service delivery as it relates to use of assets and efficiency in network design.
Numerous studies in the past 2 years have indicated that server and storage utilization rates range from 20 to 40 percent. This level of inefficiency in asset utilization has a corollary effect on power use in a given operation. Utilization can be increased in many ways that are process oriented-the larger proportion of the variable-but taking advantage of technology in the network today can have a profound complementary effect. The next section details the effect of the network on power and cooling resulting from network-based virtualization and higher utilization of infrastructure.
Energy-Efficient Data Center Solutions
An assumption in designing the infrastructure for a data center is that the fewer the power supplies in a data center, the fewer the points of conversion loss. Any time power traverses a power supply in the data center, whether it is an SMPS or a large UPS system, it loses some of its capacity in the form of heat. This loss can be generally quantified based on the manufacturer's power supply efficiency specifications. Examples can be found in most of Cisco's larger switching platforms, which have supplies that operate at 90 percent efficiency. If one of these switches requires 1000 watts (W) to operate, it can be assumed that 100W will be lost through conversion as heat: a loss that in turn adds a burden factor to cooling.
Applying this assumption throughout the planning process gives the user an efficiency comparison of not just heterogeneous systems, but systems across all IT infrastructures. With many infrastructure systems (that is, server, storage, and network) in the data center now achieving parity in the services they can support or deliver, this consideration is important when deciding where services should reside.
If fewer power supplies and superior or similar service levels are assumed, virtualization of computing, storage, and networked resources on a common platform becomes even more attractive. Virtualization in particular has the potential for a large-scale effect by increasing utilization rates for assets, thereby getting more from less.
Storage Virtualization Through Inter-VSAN Routing
Cisco provides virtualization solutions for storage infrastructure, similar to what VMware provides for servers. Using Inter-VSAN Routing (IVR), Cisco MDS 9000 family and Cisco Nexus™ products enable consolidation of disparate storage infrastructures (Figure 14).
Figure 14. Using a Common Physical Fabric to Increase Storage utilization Through IVR
In the case of Cisco IT, storage utilization was increased from 36 to 68 percent as storage capacity grew from 3 to 10 Petabytes of storage capacity. This increase resulted in a savings of US$18 million over a 3-year period in electrical costs and new-unit-acquisition deferment. More important, it allowed Cisco storage administrators to set a utilization policy that helps ensure that assets are utilized at 60 percent capacity or higher in the first year of operation. These utilization rates are confirmed by performing regular utilization audits as discussed in the "Data Center Best Practices" section of this document.
Many factors need to be considered in implementing storage virtualization, particularly its effect on facilities; the following discussion summarizes the solution.
• What is the solution?
– Storage virtualization technology using unified fabrics and IVR with the Cisco Catalyst® and Cisco Nexus platforms
• Where is the solution applied?
– Data center storage environments
• Why is the solution compelling?
– Improving storage utilization enables consolidation and allows better operating efficiency policies to be set
• How is the solution deployed?
– Through Cisco partners using the Cisco Catalyst and Cisco Nexus platforms
Figure 15 shows analysis of three use cases for security and application delivery services for a service module architecture compared to an appliance-based architecture:
• Use case 1: Isolated storage traffic for three business units sharing a common backup
• Use case 2: Combined mainframe traffic and open systems storage network connectivity in the same data center
• Use case 3: Extended storage for remote replication and tape vaulting
Figure 15. Use Case Analysis of Security and Application Delivery Services with a Service Module Architecture Compared to Appliance-Based Architecture
Source: Environmental, Social, and Governance (ESG) white paper: Building Power-Efficient Solutions with Cisco MDS 9000 Directors
Data Center Service Density Using Service Modules
Another benefit of virtualized services over the network is the consolidation of functions found in standalone appliances in the network. As more intelligence and feature sets are brought into a service module form factor, the less power per work unit performed is required-again, doing more with less.
• What is the solution?
– Application delivery and security services delivered by service modules instead of appliances using the Cisco ACE Application Control Engine Module and Cisco Catalyst 6500 Series Firewall Services Module
• Where is the solution applied?
– Data center server environments
• Why is the solution compelling?
– Providing services through service modules and at a much broader scale than appliances enables significant reductions in operating electrical overhead related to logical server groups (Figure 16)
• How is the solution deployed?
– Through Cisco partners using the Cisco Catalyst platform with service modules
Figure 16. Deploying Four Service Modules to Support Up to 250 Logical Server Groups Requires a One-Time Incremental Loading of 800W Compared to Up to 1200W per Individual Server Group Using Appliances
Optimization Points
Exploring the network as a platform for services introduces many new opportunities to increase capacity, increase efficiency, and slow power consumption growth. How applications are delivered and stored with power in mind at the enterprise level is a topic worth examining; however, this process may involve a complete reevaluation of how the current IT architecture supports the business.
If an organization is serious about stabilizing or reducing emissions growth through the strategic application of IT, then it must consider all technologies. In the case of the network, given its ability to touch anything that consumes power, fixed or mobile, more points of optimization will be identified than for any other system.
Examination of most business operations indicates that the 80:20 rule applies to power consumption efficiency. Although components that have low efficiency still exist, roughly 80 percent of the losses are from process and 20 percent are from product design. Thus, to better plan processes, planners need the right data to justify a change to the present system. The network is a tool that can help provide this data.
Many management interfaces that assist in the analysis of data gathered through the network are available today. Cisco and its partners offer solutions that enable users to identify focus areas for service optimization. A good way to start this analysis is to examine how a given operation delivers an application, with the understanding that every time a packet is routed, switched, cracked, served, or stored, an associated power footprint is created. A power footprint corresponds to a carbon footprint. Use of the right mix of management applications allows a user to identify the right areas to focus on to achieve carbon neutrality.
Migration Strategies
After IT has a baseline understanding of the fundamentals of data center power efficiency and has identified optimization points, a migration strategy can be developed. For the sake of this document, the implementation of this strategy focuses on the operating efficiency of the data center and may include the physical migration of assets to different environments. The scope for a given enterprise would, of course, be user-defined, but the following sections provide supporting data and best practices that can help planners formulate strategies concerning the network. It is also important to note that the goal of migrating toward a carbon-neutral operation is a perpetual effort.
Users often supplement internal skill sets by engaging with professional service organizations, particularly when establishing an efficiency baseline for the data center operation, which can be a complex and labor-intensive process. However, establishment of a baseline is necessary to migrate to and improve the operating efficiency of the data center.
As mentioned earlier, calculating efficiency includes calculating the total power consumed per system. However, this infrastructure assessment needs to be with the services the data center offers. For example, of the systems that are calculated, what portion supports security, access, application delivery, and so on? This analysis provides a power profile per service or application, enabling before-and-after analysis.
When a service power profile analysis is complete, more accurate decisions can be made regarding changes in the infrastructure architecture. The analysis should include facilities as well as IT infrastructure; then a sound migration strategy can be developed. The migration is from lower efficiency and utilization to better efficiency and utilization, and it may or may not include a physical relocation of assets.
Efficiency-Assessment Services
Cisco Data Center Efficiency Services can help support a customer's green agenda by providing efficiency benchmarks. These benchmarks, based on standards established in part by The Green Grid, provide metrics for success in building the business case to improve operating efficiency for the data center and supporting operations.
These services help enterprise data center managers across both facilities and IT take control of their power growth and costs and the related green implications. This assessment is accomplished through a standard methodology focused on the following:
• Identify how much power and cooling is being used and where
• Establish an efficiency benchmark for the current data center power and cooling design
• Design and implement more efficient data center architectures
With a strong collaborative partnership with APC and Emerson, these services are being rolled out globally, including a trio of new services:
• Cisco Facilities Assessment Services
• Cisco Energy Efficiency Assessment Services
• Cisco Architecture Assessment Service
The Cisco Data Center Efficiency Services are part of the Cisco Efficiency Assurance Program. This web-based portal helps Cisco customers better analyze the business case for green technologies through planning tools, a learning center, and these new services, which is the first step in any green transformation (Figure 17).
Figure 17. Efficiency Assessment Services Portfolio Provide an Incremental Methodology to Benchmark and Improve Energy Efficiency
In addition, Cisco Advanced Services for the data center provide a user-definable scope, allowing Cisco service teams to work in partnership with customers to develop the best approach to meet business priorities, which allows Cisco to focus on not just data centers but IP-based energy management and the facilities that support IT, for example.
Data Center Best Practices
This document discusses various aspects of designing efficiency into a data center operation using Cisco and partner technologies. In addition to the solutions highlighted, many best practices were identified. Table 1 provides a consolidated overview of these best practices.
Table 2. Best Practices Being implemented By Cisco IT to Improve Energy Operating Efficiency
Frequently Asked Questions
Q. How is Cisco Efficiency Assurance Program green?
A. Green programs generally cover reduce, reuse, and recycling of finite resources. While Cisco has mature e-waste and recycling programs, it has not formally addressed energy efficiency to date. Energy efficiency is one of three main areas that can combat climate change, the other two being production of cleaner power and carbon sequestration. The Cisco Efficiency Assurance Program is green because it can help you better plan and eventually manage how you use what is today a finite resource: energy. Seventy percent of the world electricity still comes from the burning of fossil fuels, and almost 100 percent of the energy used for travel comes from fossil fuels, so anything that can be done to use energy more efficiently will have a positive effect on reduction of CO2 emissions.
Q. How does the Cisco Efficiency Assurance Program help me be more green as a professional?
A. One of the biggest hurdles in adopting a green mindset is learning a whole new system of choices. Many people want to be involved, but still relatively few know what to do. The Cisco Efficiency Assurance Program is a central interface for collaboration between Cisco and a wide range of users on adopting a sustainable IT strategy. In time, the program will provide extensive wiki resources to facilitate this collaboration; today, a feedback loop sends comments directly to the program's chief designer.
Q. What is "sustainability" and how does it differ from "green"?
A. Sustainability is a more precise and scientific term than green, which has broader social and political implications. Both connote a need for humans to live more in balance with the earth's natural systems and carrying capacity for life. Adopting sustainability as an organizing principle for change in systems narrows the scope of considerations. A concise scope helps ensure that the economic case can also be assessed business decisions involving sustainable transformation.
Q. Why now?
A. As a publicly traded company, Cisco is obligated to its shareholders and, in turn, its customers to demonstrate value. Energy efficiency in IT is a relatively new focus area, emerging about the same time that blade server form factors came to market, around 2005. Most of the attention to date on energy consumption has focused on servers because servers are the number-one consumer of power in data centers. In general, servers account for about 50 percent of power consumption, storage 35 percent, and the network 15 percent. Given the network's much smaller power footprint, Cisco's customers have only recently begun to raise concerns about the power allocated to networks and the corresponding emissions. Conversely, server and, given the current high growth rates, storage vendors have been getting this feedback for a longer time period. Cisco has always maintained a progressive attitude toward conservation of resources as evidenced in Cisco's annual Citizenship Report. This report shows how Cisco endeavors to reduce its footprint as a user of resources. Helping customers reduce their carbon footprint through the strategic integration of Cisco technologies is a more recent endeavor.
Q. Where are the future plans for the Cisco Efficiency Assurance Program?
A. In its first iteration, the Cisco Efficiency Assurance Program is focusing primarily on two goals:
• Centralization of product and technology content that will help customers reduce their footprint
• The first level of interpretation of energy and carbon, primarily related to asset utilization
With its many user communities in mind, Cisco is seeking to develop this program in a democratic fashion. Said another way, Cisco is involving the human network in the shaping of this tool to help ensure that it will be relevant, useful, and timely. Current plans for future versions of the program include continuation of the planning tool approach shown in the programs beta release.
Another area identified under planning focuses on site selection. One of the greenest actions an organization can take is to not power a business operation using high carbon fuel sources. Locating a high-power-consumption operations such as a data centers in a geographical area with cleaner power can have a significant positive effect. Cisco is currently investigating how to make the search for green power more accessible to business planners.