Intel® Cloud Builders Guide to Cloud Design and Deployment ... · Executive Summary The evolution of cloud computing has resulted in highly efficient and carefully optimized data

Intel® Cloud Builders Guide to Cloud Design and Deployment on Intel® PlatformsManage Data Center Carbon Footprint with Dell, Intel, and JouleX

Audience and PurposeThis reference architecture outlines the usage of data center energy management technologies to monitor, report, and manage carbon footprint in cloud data centers. This document addresses the key aspects of the usage model Carbon Footprint Values as defined by Open Data Center Alliance1 and illustrates the accurate reporting of carbon footprint with real time instrumentation of power consumption data from the server platforms enabled by technologies and solutions from Dell, Intel, and JouleX. The technologies and solutions used in the guide are the same as that in “Data Center Energy Management with Dell and JouleX” published by Intel® Cloud Builders Program2. This paper should be used along with the above mentioned documents. The techniques and results described can be used as a reference to understand Carbon Footprint management in data centers with the use of hardware and software components presented using JouleX* Energy Management Solution and Dell PowerEdge* C-Series Servers implementing Intel® Power Management technologies.

September 2011

Intel® Cloud Builders GuideIntel® Xeon® Processor-based ServersManage Data Center Carbon Footprint with Dell, Intel, and JouleX

Intel® Xeon® Processor 5500 Series

Intel® Xeon® Processor 5600 Series

2

Intel® Cloud Builders Guide: Manage Data Center Carbon Footprint with Dell, Intel, and JouleX

Table of Contents

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Cloud Computing and Carbon Footprint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Government Regulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Energy Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

JouleX Energy Management Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Dell PowerEdge* C-Series Servers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Intel® Power Management Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Test-bed Blueprint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Design considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Software Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Hardware and Software Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Physical Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Setup and Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Carbon Footprint Use Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Carbon Emission Calculation and Usage Based on ODCA Usage Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Carbon Footprint Monitoring and Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Carbon Footprint Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Things to consider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Executive SummaryThe evolution of cloud computing has resulted in highly efficient and carefully optimized data centers with increased server density and capacity that makes considerations on carbon footprint and energy consumption extremely critical along with several other factors that were not as significant in smaller data centers of the past. Increasing government regulations and guidelines on carbon footprint along with social responsibility awareness by companies are driving data center operators to monitor, report, and control the carbon footprint of their data centers. To support this evolution, Intel works with end users to create an open data center roadmap of usage models that address requirements like this and other key IT pain points for more secure, efficient, simple and greener cloud architectures built on a foundation of transparency.

This paper describes a carbon footprint reference architecture based on Dell, Intel, and JouleX solutions for monitoring, reporting, and managing carbon footprints. The solutions described in this document address key aspects of the usage model Carbon Footprint Values as defined by Open Data Center Alliance1 (ODCA) and uses technologies and solutions explained in the guide “Data Center Energy Management with Dell and JouleX” published as part of the Intel® Cloud Builders Program2.

In this reference architecture we used Dell PowerEdge* C-Series Servers3 with Intel® Intelligent Power Node Manager4 (Intel® Node Manager) and JouleX Energy Management Software5 which uses Intel® Data Center Manager6 (Intel® DCM). These technologies enable reporting of actual carbon footprints of the data centers with power consumption instrumented real time from the server platforms, as against the methods based on estimation and modeling of energy consumption.

This would result in more accurate carbon footprint reporting and carbon tax payment if applicable.

We illustrate the following carbon footprint use cases.

1. Carbon emission calculation and usage based on Open Data Center Alliance Usage Models illustrates how CO2 emission values are calculated and used for carbon footprint monitoring and reporting in data centers.

2. Carbon footprint monitoring and reporting at data centers. In this use case we illustrate how the server platforms like Dell PowerEdge* C-Series servers help enable accurate monitoring and reporting of a carbon footprint with real time and accurate instrumentation of power consumption of server platforms by Intel Node Manager technology as against the traditional methods of estimation and modeling.

We also explain the power management usage models needed to optimize planning and data center power capacity as well as energy consumption, which would in turn result in a carbon footprint reduction. It is important to note the optimization methods and solutions are still evolving and not all of them are fully automated, as that requires design and integration of multiple data center management components and solutions.

Cloud Computing and Carbon FootprintBased on the Environmental Protection Agency’s report to the government, in 2006 data centers in the US consumed about 1.5 percent of the nation’s energy and were poised to double this by 20117.

According to Gartner estimate in 2007, Information and Communications Technology (ICT) accounts for 2 percent of global CO2 emissions, placing it on par

with the aviation industry8. And data centers account for 23 percent of global ICT CO2 Emissions.

According to market research and consulting firm Pike Research, data centers around the world consumed 201.8 terawatt hours (TWh) in 2010 and energy expenditures reached $23.3 billion. That’s enough electricity to power 19 million average U.S. households. The good news is that, according to Pike Research, the adoption of cloud computing could lead to a 38 percent reduction in worldwide data center energy expenditures by 20201.

Cloud computing is the new model for IT services that has emerged to break the trend of decline in flexibility combined with increase in costs. It is an approach to computing that uses the efficient pooling of an on-demand, self-managed infrastructure, consumed as a service. This approach separates applications and information from the complexity of underlying infrastructure, so IT can support and enable business value.

While server performance-per-watt continues to increase, the energy consumed per server also continues to rise. These advancements enable increasing number of servers and density in modern data centers, making planning and management of power and cooling resources critically important to ensure efficient utilization of provisioned capacity. In order to realize the vision of cloud computing, new technologies are needed to address power efficiency and energy management. These will become fundamental to architectures from the micro-processor stage up through the application stack.

The server consolidation, operation, and design efficiency improvements provide opportunity to monitor and report carbon footprint and energy usage, and take effective actions to control it.

3


Government RegulationsAs data center footprints expand and energy consumption rises, organizations are under increasing pressure to report and reduce their environmental impact, as well as reduce the cost of their operations. In many regions of the world, governments and corporations now require a measurement of the carbon footprint of services and products.

For example, the Carbon Disclosure Project9 is persuading businesses to issue Corporate Responsibility Reports which include carbon reporting. Reducing the amount of electrical energy wasted in business operations can be a win-win proposition for both businesses and the planet; but doing so at the data center level requires an understanding of how technology platforms consume electrical resources.

The Australian government has unveiled plans to impose a tax on carbon emissions for the worst polluters10. Under the new scheme set to begin on 1 July 2012, the government plans to include any company that produces at least 25,000 tons of carbon dioxide per year with some exceptions at a tax rate of A$23 ($25) per tonne.

The CRC Energy Efficiency Scheme in UK that came into effect in April 2010 is a mandatory carbon emissions reporting and pricing scheme to cover all organizations using more than 6,000MWh per year of electricity (equivalent to an annual electricity bill of about £500,000)11.

Finland, Sweden, Great Britain, New Zealand, Boulder (Colorado, USA), Quebec (Canada), British Columbia (Canada) are the other regions that have implemented

Figure 1: Estimated Carbon Dioxide Emission Rate from US Electric Plants

or considering carbon taxes in various forms12.

Energy SourcesThe carbon footprint in a data center would be calculated based on energy sources and data center power efficiency, which will be explained later in this document. Energy cost and CO2 emissions from energy sources are important considerations when determining data center locations for large cloud data centers today.

Plenty of information is available publicly on the CO2 emissions rates. For example, Figure 1 below shows the carbon dioxide emission rate from generating units at US electric plants13.

4


JouleX Energy Management Solutions The JouleX Energy Manager (JEM) reduces energy costs and carbon footprints in data centers by monitoring, analyzing, and controlling energy usage of servers and network-connected devices and systems with no client-side agents, stubs or hardware required. Unlike other products, JEM is a single solution that provides a global view of energy consumption for devices such as PCs, servers, VoIP phones and devices, printers, network switching devices, IP power switches, HVAC systems, and more.

JEM for the Data Center

JEM enables Load Adaptive* Energy Management for the data center, based on the principles of conservation and optimization. Use JEM to allocate the right amount of power only to those devices that need to perform productive work, and minimize the energy supplied when idle or operating at less than full capacity. JEM accurately measures system and application utilization and energy loads at the server, virtual machine, and application level while dynamically allocating computing resources as they are needed. JEM pinpoints underutilized and low density servers consuming the most energy, which are prime candidates for virtualization. JEM also identifies dead and idle servers for retirement. JEM’s overall approach incorporates Load Adaptive Computing and Networking to help reduce data center energy costs by as much as 60 percent.

JEM for the Enterprise

Implementing green business initiatives has become a top priority for socially and environmentally conscious corporations that want to stay ahead of the regulatory

curve. A sustainable business is also important to the financial bottom line. In most companies, energy consumption presents the largest opportunity to impact the environment and save money. JEM gives you the ability to determine a baseline and monitor, analyze, and manage energy consumption across the enterprise. This intelligence can be used to:

• Measure energy utilization/consumption and carbon emissions by device

• Support sustainable procurement initiatives with actual energy consumption metrics

• Create enterprise sustainability reporting for a variety of metrics, including energy consumption by device, energy savings, carbon savings, and more

• Create and implement event-based policy, rule-based policy, energy usage simulations, and ROI modeling

JEM’s policy-based energy optimization capabilities enable you to:

• Power manage distributed office equipment

• Optimize virtualization and cloud computing energy in the data center

• Provide automated demand response

• Perform Load-Adaptive Computing and Networking that allocates the right amount of power only to those devices that need to perform productive work

Dell PowerEdge* C-Series Servers Dell has extended its PowerEdge* server family with the new C-Series. Designed with inspiration from Dell’s Data Center Solutions (DCS) business, these new servers are optimized for performance and efficiency for scale-out customers in HPC, Web 2.0, hosting, gaming, and public and private cloud builders. The PowerEdge C servers include:

• PowerEdge* C5220 Microserver: Enhance your hosting, content delivery network (CDN) or Web 2.0 performance, cooling and efficiency with up to 12 high-density, feature-rich PowerEdge C5220 microservers in a PowerEdge* C5000 shared-infrastructure chassis. Intel® Xeon® E3-1200 processor with 2 or 4 cores supports up to 65W thermal design power (TDP) for the 12-sled microserver and up to 95W TDP for the 8-sled version

• PowerEdge* C1100: increased-memory, power-efficient, cluster-optimized compute node server (1U/2S, Up to 192GB RAM, Intel® Xeon® processor 5500/5600 series, 2 x 1GbE Intel 82576 Kawela ports). Great for power and space sensitive customers requiring maximum memory flexibility.

• PowerEdge* C2100: high performance data analytics, cloud compute platform and cloud storage server (2U/2S, Up to 192GB RAM, Intel Xeon processor 5500/5600 series, 2 x 1GbE Intel 82576 Kawela ports) Great for scale-out data center environments where memory and storage density matter most: Hadoop, Map/Reduce, Web analytics, database.

• PowerEdge* C6100: 4-node cloud and cluster optimized shared infrastructure server (2U/Up to 4 2S server nodes [hot-serviceable], Intel Xeon processor 5500/5600). Great for Hyperscale-inspired building block for high-performance cluster computing (HPCC), Web 2.0 environments, and cloud builders where the performance is key.

Intel® Power Management TechnologiesMicroprocessors are possibly the most energy intensive components in servers and have traditionally been the focus of power management strategies. Emergent technologies such as solid state drives

5


have the potential to significantly reduce power consumption and in the future, management of memory power consumption may be incorporated.

Intel Node Manager and Intel DCM are designed to address typical data center power requirements such as described above.

Intel Node Manager is implemented on Intel server chipsets starting with Intel Xeon processor 5500 series platforms onwards. Intel Node Manager provides power and thermal monitoring and policy based power management for an individual server and is exposed through a standards-based IPMI interface14 on supported Baseboard Management Controllers (BMCs). Node Manager requires an instrumented power supply that conforms to the PMBus standard15

Intel DCM SDK provides power and thermal monitoring and management for servers, racks, and groups of servers in data centers. Management Console Vendors (ISVs) and System Integrators

(SIs) can integrate Intel DCM into their console or command-line applications to provide high value power management features. These technologies enable new power management paradigms and minimize workload performance impact.

Intel Intelligent Power Node Manager

Intel Xeon processors regulate power consumption through voltage and clock frequency scaling. Reduction of the clock frequency reduces power consumption, as does lowering voltage. The scale of reduction is accomplished through a series of discrete steps, each with a specific voltage and frequency. The Intel Xeon processor 5500 series can support 13 power steps. These steps are defined under the ACPI16 standard and are colloquially called P-states. P0 is nominally the normal operating state with no power constraints. P1, P2, and so on aggressively increase the power capped states.

Voltage and frequency scaling also impacts overall system performance, and therefore will constrain applications. The

Figure 2: Intel Node Manager Power Management Closed Control Loop

control range is limited to a few tens of watts per individual microprocessor. This may seem insignificant at the individual microprocessor level, however, when applied to thousands or tens of thousands of microprocessors typically found in a large data center, potential power savings amount to hundreds of kilowatt hours per month. Intel Node Manager is a chipset extension to the Baseboard Management Controller (BMC) that supports in-band/out-of-band power monitoring and management at the node (server) level. Some of the key features include:

• Real-time power monitoring

• Platform (server) power capping

• Power threshold alerts

Figure 2 below shows the Intel Node Manager server power management closed control loop.

6


Intel Data Center Manager

Intel DCM SDK enables ISVs to provide high value power management features that address power and thermal issues challenging IT organizations.

• Monitor power and thermals – Aggregated actual and historical trend data and alerts for racks and groups of servers

• Policy-based management – Intelligent heuristics engine maintains group power cap on demand

• Scalability – Manage thousands of nodes using agentless technology

• Robust implementation – Reference GUI and extensive validation with data center proof-of concepts to lower deployment risks

Test-bed Blueprint Intel has worked with Dell and JouleX to implement a test bed that features Dell PowerEdge C-Series servers which are designed specifically for power and space

sensitive data centers. The test bed is intended to provide a flexible environment to simulate those aspects of a commercial data center that are relevant to cloud computing usage models. JouleX Energy Management software uses Intel DCM as an integrated component.

Design Considerations

Intel Node Manager compliant servers along with PMBus power supply for real-time power monitoring are required.

Software Architecture

The following diagram shows a high level view of JEM components.

Two of the components interact directly with the network infrastructure: Asset Connectors and Device Proxies.

Asset Connectors connect to the network to discover and import existing devices into JEM. For instance, using Asset Connectors, JEM can import all devices from Active Directory. Device Proxies work hand in hand with Asset Connectors to implement low-level communication

protocols with existing devices, like WMI, to communicate with Microsoft Windows*servers, SSL for Linux*, SNMP for networking equipment like switched and routers and many more.

For power management, Intel DCM is one of the Asset Connectors used in JEM using IPMI, DCMI, or other communication protocols supported by Intel DCM to monitor and manage power at real time.

The figure below shows how Intel DCM works in JEM along with other supported capabilities of JEM. Intel DCM provides power monitoring and management of supported devices including Dell PowerEdge C Series servers using out of band communication, while JEM uses in band communication with the appropriate Device Proxies to get other asset and utilization information. Power data along with other information provides compelling reporting and management capabilities.

Figure 3: JouleX Energy Management components

7


Figure 4: JouleX Energy Manager with Intel Data Center Manager

8


Hardware and Software Description

JouleX Energy Management Server

Virtual Machine hosted on VMWare

4 CPUs, 6 GM RAM, 50GB hard disk

Software Microsoft Windows 2008 R2 64 bit, .NET 4.0Intel DCM 2.3 or laterJEM software 2.6 or later

Server 1 Dell PowerEdge C1100 2-way Intel Xeon Processor E5570 @ 2.93GHz with 12GB RAM, 250GB SATA HDD

Intel Node Manager enabledBMC CardPMBus power supply

Software CentOS Release 5.5

Server 2 Dell PowerEdge C2100 2-way Intel Xeon Processor E5620 @ 2.40GHz with 12GB RAM, 500GB SATA HDD


Software Windows 2008 R2 64 bit, SQL Server 2005 workload

Server 3 & 4 Dell PowerEdge C6100 2-way Intel Xeon Processor E5530 @ 2.40GHz with 12GB RAM, 250GB SATA HDD


Software Windows 2008 R2 64 bit, SQL Server 2005 Workload

Server 5 to 16 Dell PowerEdge® C5220 One Intel Xeon Processor E31280 @ 3.5GHz with 4GB RAM, 500GB SATA HDD


Software Windows 2008 R2 64 bit, SQL Server 2005 Workload

Table 1: Hardware description

9


Physical Architecture

Figure 5 shows the test bed deployment architecture. JEM and DNS/DHCP services are installed on virtual machines. The four Dell server nodes are used for use case testing with one node each from Dell PowerEdge C1100 and Dell PowerEdgeC2100 systems, two nodes from Dell and twelve nodes from Dell PowerEdgeC5220 system. These systems have Intel Node Manager implemented. JEM connects to the system in-band to monitor and collect host information and

Figure 5: Physical Architecture of Test Bed Setup

out of band via Intel DCM to monitor and manage power consumption.

Setup and Configuration

The hardware and software set up instructions are explained in detail in the document, "Intel® Cloud Builders Guide: Cloud Design and Deployment on Intel® Platforms: Data Center Energy Management with Dell, Intel, and JouleX". Please refer to that document along with this paper.

Carbon Footprint Use Cases

Carbon Emission Calculation and Usage Based on ODCA Usage Models

The calculation and configuration of CO2 Emission values would be based on ODCA usage model on carbon footprint. We recommend reading the ODCA document for detailed understanding and considerations. In this reference architecture we will address the usage model with Dell PowerEdge C-Series servers implementing Intel Node Manager.

10


However the methodology can be extended to other servers and devices as well.

It should be understood that two related concepts – energy efficiency and carbon footprint - are in consideration here, which may be reflected respectively in the costs and climate impact of services.

• Energy efficiency is to be encouraged, but that factor is addressed elsewhere by ODCA, especially in the cost of services.

• The carbon footprint assessment described herein deals with the key indicator of carbon emission.

Carbon emissions of electricity source(s)[ [Energy overhead

of data center[ [kWh Energy Used per Unit[ [Amount of IT

equipment used in standard units

[ [

The two indicators are not synonymous. When a data center uses green or carbon-neutral energy from a wind turbine or nuclear power, the carbon footprint is virtually zero. This does not mean though that the cloud provider has an energy-efficient data center.

As for carbon offsetting, there is some debate as to whether it is acceptable or not. The method we’re providing here therefore allows for a number of values to be measured: the energy used, what proportion of that is renewable, and whether the residue is offset or not. Transparency is required on the part of

the cloud provider for the cloud subscriber to be satisfied that they are receiving valid figures.

The method of assessing the carbon footprint depends on obtaining and combining data from a number of internal and external sources. There is, as yet, no one globally accepted and universally applicable methodology, but methods are in development and there are well understood and agreed upon approaches.

The amount of carbon produced can be derived from the following formula as per the ODCA Usage Model.

The energy overhead of a data center is usually expressed a Power Usage Effectiveness (PUE), as defined by the Green Grid17.

As per the ODCA Usage Model guideline, the electricity used per unit should not just be a “nameplate” value, as this is a maximum figure that does not indicate a real figure for normal use. For servers, the amount of electricity used tends to be around 45–50 percent of the nameplate value. Many of the data centers use this estimation model today to report carbon footprint; but even this tends to be conservative and could result in over estimation and additional carbon tax payments if applicable.

The platforms implementing Intel Node Manager, like Dell PowerEdge C-Series servers, have the ability to report real time power consumption of the servers, which would result in a more accurate calculation and reporting of the actual carbon footprint in data centers.

In JEM, the power consumption of the server is accurately monitored in real time with the power consumption directly instrumented from Dell PowerEdge server platforms enabled by Intel Node Manager. JEM has the capability of configuring CO2 emissions by location that will be automatically applied to the servers assigned to that location. So in the above formula we need only CO2 emissions from energy sources and energy overhead of data center (PUE). JEM will generate reports based on this.

Steps for configuring CO2 emission values in JEM is explain below.

11


Configure CO2 emission values in JEM

JEM allows setting energy costs and carbon emissions depending on the location and source of energy for data centers. These values would be used for calculations while generating the carbon footprint reports.

For example, if the CO2 emission from the energy source is 0.6 kg/KWh and energy overhead of the data center (PUE) is 1.5, CO2 emission value = 0.6 x 1.5 = 0.9 kg/KWh

• On JEM console, go to Settings>Energy Prices.

CO2 emission values can be added here. In this case we added 0.9 kg/KWh to the San Francisco location.

12


• Now let us check the servers in this location. Go to Devices>Location>San Francisco.

• Double-click on one of the servers, then click ‘Summary’, and select CO2 rate from the drop down menu. The CO2 rate for this server is 0.9 kg/kWh. This rate will be applied for the carbon emission reports for this server.

13


Carbon Footprint Monitoring and ReportingAs explained in the previous section, the key for accurate reporting of a carbon footprint is the actual measurement of power consumption in real time. This is enabled by Intel Node Manager and implemented in Dell PowerEdge C-Series

server platforms. With the real time aggregation of this data and computation of the carbon footprint reports by JEM with integrated Intel DCM rich, accurate and actionable carbon footprint reports can be generated with server level granularity and different levels of grouping and aggregation.

Thus the power management solutions from Dell, Intel, and JouleX help to generate actual reporting of a carbon footprint and payment of carbon taxes as required, not the typical high figures generated with estimation.

To generate CO2 emission reports, go to Reports>Detailed Reports>Carbon Emissions

a. A carbon emission report by ‘Business Unit’ is shown below.

14


b. A carbon emission report by ‘Location’ is shown below.

15


c. The following report shows the total CO2 emission of all devices.

JEM has rich reporting capabilities. Reports can be filtered further by any metadata of the servers like location, device type, model, operating system, specific server name, etc. The report period can also be set for week, month, etc. This allows the user to generate reports at different levels of granularity, allowing comparing, analyzing and taking appropriate optimization options.

16


Carbon Footprint ManagementApart from monitoring and reporting, there are other use cases and techniques that can be used to optimize data center energy consumption and thus reduce energy cost and carbon footprint. The use cases are explained in another document using the same solutions and technologies covered in this paper. Please refer to the document, "Data Center Energy Management with Dell and JouleX" published at Intel Cloud Builders Program.

The use cases are summarized below. It is important to note the optimization methods and solutions are still evolving and not all of them are fully automated as it requires design and integration of multiple data center management components and solutions.

1. Real-time Server Energy Usage Monitoring, Reporting & Analysis to get continuous and actual energy usage visibility via agentless monitoring of the servers along with other devices and systems in the enterprise network, data center and facilities. The actionable reporting and analysis with real time power monitoring enables reduction in energy cost and carbon emissions.

2. Power Guard Rail & Optimization of Rack Density by imposing power guard to prevent server power consumption from straying beyond preset limit. The deterministic power limit and guaranteed server power consumption ceiling helps maximize server count per rack and therefore return of investment of capital expenditure per available rack power when rack is under power budget with negligible or no per server performance impact.

3. Disaster Recovery/Business Continuity by applying significantly lower power caps to lower power consumption and heat generation

when unforeseen circumstances like power outage and cooling system failure occurs. In these scenarios it may be appropriate to set aggressively lower power caps too, though performance would be affected. The use case illustrates how it works at a data center location or a group of servers.

4. Power Optimized Workloads to achieve power efficiency. Workload profiles are built and a maximum performance loss target set. Experiments determine how much capping can be applied before the performance target is hit. The approach is to match actual performance against service level requirements. For workloads that were not processor intensive, we were able to optimize server power consumption by approximately 20 percent without an impact on performance. For workloads that were processor intensive, for the same 20 percent power savings, we saw an 18 percent decrease in performance. For a 10 percent power reduction, performance decreased by 14 percent.

5. Data Center Energy Reduction through Power Aware Support for Multiple Service Classes showcases the ability to enforce multiple SLAs across different populations of users with different priority workloads. Workloads that ran over a period of eight hours realized 25 percent less energy consumption.

Things to ConsiderPlease refer to the document “Data Center Energy Management with Dell and JouleX” published by the Intel Cloud Builders Program for things to consider like architecture and scalability while deploying the solutions explained in this paper.

GlossaryIntel® Intelligent Power Node Manager:

Intel® Intelligent Power Node Manager (Intel® Node Manager) resides on Intel® Xeon® 5500 server (and later) platforms. It provides power and thermal monitoring, and policy based power management for an individual server. Capabilities are exposed through standard IPMI interface from supported Baseboard Management Controllers (BMC). This requires an instrumented power supply such as PMBus*.

Intel® Data Center Manager:

Intel® Data Center Manager (Intel DCM) scales Intel Node Manager functions to racks and groups of servers and enables IT users to benefit from increased rack density, reduced capital, and operational expenses

JouleX Energy Management Software:

The JouleX Energy Manager (JEM) reduces energy costs by monitoring, analyzing, and managing energy usage of all network connected devices and systems, without the use of costly and unwieldy software agents.

17


References1. Open Data Center Alliance Usage:

Carbon Foot Print Values, http://www.opendatacenteralliance.org/document-sections/category/71-docs?download=434%3Acarbon-footprint-values

2. Intel® Cloud Builders Guide: Cloud Design and Deployment on Intel® Platforms: Data Center Energy Management with Dell, Intel, and JouleX, http://www.intelcloudbuilders.com/library?view=displaysearch&filter_companyname=27&company_name=joulex

3. Dell PowerEdge C Series, http://www.dell.com/content/topics/topic.aspx/global/products/landing/en/poweredge-c-series?c=us&l=en&s=gen&redirect=1

4. Intel® Intelligent Power Node Manager, http://www.intel.com/content/www/us/en/data-center/data-center-management/intelligent-power-node-manager-general.html

5. JouleX Energy Management Software, http://www.joulex.net/products-0/joulex-energy-manager-for-data-centers/

6. Intel® Data Center Manager, http://software.intel.com/sites/datacentermanager/index.php

7. EPA Report to Congress on Server and Data Center Energy Efficiency, http://www.energystar.gov/ia/partners/prod_development/downloads/EPA_Report_Exec_Summary_Final.pdf

8. Gartner Says Data Centers Account for 23 Per Cent of Global ICT CO2 Emissions, http://www.gartner.com/it/page.jsp?id=530912

9. See the Carbon Disclosure Project, https://www.cdproject.net/

10. Australia plans to impose carbon tax on worst polluters, http://www.bbc.co.uk/news/world-asia-pacific-14096750

11. Where Carbon is taxed, http://www.carbontax.org/progress/where-carbon-is-taxed

12. Carbon Dioxide Emissions from the Generation of Electric Power in the United States, http://www.nrdc.org/air/pollution/benchmarking/2008/benchmark2008.pdf

13. Intelligent Platform Management Interface, http://www.intel.com/design/servers/ipmi/ipmi.htm

14. PMBus, http://pmbus.org/specs.html

15. Advanced Configuration & Power Interface, http://www.acpi.info

16. See Green Grid data center power efficiency metrics: PUE and DCiE, http://www.thegreengrid.org

17. CRC Energy Efficiency Scheme in England starting in 2010, http://www.carbontrust.co.uk/policy-legislation/business-public-sector/pages/carbon-reduction-commitment.aspx

For More Information

• Intel® Cloud Builders: www.intel.com/cloudbuiders

• Open Data Center Alliance: www.opendatacenteralliance.org

• Dell: www.dell.com/cloud

• JouleX: www.joulex.net

18


http://www.opendatacenteralliance.org/document-sections/category/71-docs?download=434%3Acarbon-footprint-values





http://www.intelcloudbuilders.com/library?view=displaysearch&filter_companyname=27&company_name=joulex




http://www.dell.com/content/topics/topic.aspx/global/products/landing/en/poweredge-c-series?c=us&l=en





http://www.intel.com/content/www/us/en/data-center/data-center-management/intelligent-power-node-manager-general.html




http://www.joulex.net/products-0/joulex-energy-manager-for-data-centers/



http://software.intel.com/sites/datacentermanager/

http://software.intel.com/sites/datacentermanager/

http://www.energystar.gov/ia/partners/prod_development/downloads/EPA_Report_Exec_Summary_Final.pdf




http://www.gartner.com/it/page.jsp?id=530912

http://www.gartner.com/it/page.jsp?id=530912

https://www.cdproject.net/en-US/Pages/HomePage.aspx

http://www.bbc.co.uk/news/world-asia-pacific-14096750

http://www.bbc.co.uk/news/world-asia-pacific-14096750

http://www.carbontax.org/progress/where-carbon-is-taxed/



http://www.nrdc.org/air/pollution/benchmarking/2008/benchmark2008.pdf



http://www.intel.com/design/servers/ipmi/ipmi.htm

http://www.intel.com/design/servers/ipmi/ipmi.htm

http://pmbus.org/specs.html

http://www.acpi.info/

http://www.thegreengrid.org/

http://www.carbontrust.co.uk/policy-legislation/business-public-sector/pages/carbon-reduction-commitment.aspx




http://www.intel.com/content/www/us/en/cloud-computing/cloud-builders-provide-proven-advice.html

http://www.intel.com/content/www/us/en/cloud-computing/cloud-builders-provide-proven-advice.html

http://www.opendatacenteralliance.org/

http://www.opendatacenteralliance.org/

http://content.dell.com/us/en/enterprise/cloud-computing.aspx

http://www.joulex.net/

Disclaimers∆ Intel processor numbers are not a measure of performance. Processor numbers differentiate features within each processor family, not across different processor families. See www.intel.com/

products/processor_number for details.INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL® PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROP-

ERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. UNLESS OTHERWISE AGREED IN WRITING BY INTEL, THE INTEL PRODUCTS ARE NOT DESIGNED NOR INTENDED FOR ANY APPLICATION IN WHICH THE FAILURE OF THE INTEL PRODUCT COULD CREATE A SITUATION WHERE PERSONAL INJURY OR DEATH MAY OCCUR.

Intel may make changes to specifications and product descriptions at any time, without notice. Designers must not rely on the absence or characteristics of any features or instructions marked “reserved” or “undefined.” Intel reserves these for future definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them. The infor-mation here is subject to change without notice. Do not finalize a design with this information.

The products described in this document may contain design defects or errors known as errata which may cause the product to deviate from published specifications. Current characterized er-rata are available on request. Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order. Copies of documents which

have an order number and are referenced in this document, or other Intel literature, may be obtained by calling 1-800-548-4725, or by visiting Intel’s Web site at www.intel.com.Copyright © 2011 Intel Corporation. All rights reserved. Intel, the Intel logo, Xeon, Xeon inside, and Intel Intelligent Power Node Manager are trademarks of IntelCorporation in the U.S. and other countries. *Other names and brands may be claimed as the property of others.


Documents

Intel® Cloud Builders Guide to Cloud Design and Deployment ... · Executive Summary The evolution of cloud computing has resulted in highly efficient and carefully optimized data