Energy efficiency is an important strategic goal for data center operators for a number of reasons. Several indicators and metrics that are widely discussed today are available to help manage data centers successfully from an energy perspective.
However, energy efficiency is just one way of looking at a data center: Data center service providers, like enterprise data centers, colocation providers and others, find themselves in a challenging and fast-changing competitive environment. Because of these challenges, data center managers also need additional key performance indicators to link operations to business and to ensure they are on track regarding their overall strategies.
FNT developed a concept for Data Center Business Value Dashboards that is based on the established method of a balanced scorecard to help implement and measure strategies for successful data center operation. In order to follow such a strategic path, a high degree of transparency is required with respect to business strategy, underlying products and services, and all related infrastructure assets. A comprehensive central data model that holds all relevant information and indicators to feed a Business Value Dashboard is essential for such an approach.
The following approach is based on FNT Command and FNT ServicePlanet, FNT’s standard software products for service and infrastructure management. FNT is a leading provider of integrated software solutions for IT management, data center infrastructure management and telecommunication infrastructure management worldwide. FNT‘s innovative software suite, FNT Command, is used worldwide as an infrastructure management application for service providers, enterprises and governmental organizations by more than 25,000 users since 1994.
FNT Command is easy-to-use and provides transparency in an integrated way for the management of IT, network infrastructure, data centers and telecommunication. The deep integration of all data items and the comprehensive data model of FNT Command are unique in the software market. It builds the central resource repository for all planning, fulfillment and assurance processes within the business areas of telecommunications, cable networks, outside & inside plant management, data center infrastructure management as well as IT infrastructure management.
FNT ServicePlanet, a unique software for product portfolio and service management, provides the foundation for defining, managing, and monitoring all business services and service assets over the entire service lifecycle. Standardizing products and their components in FNT ServicePlanet allows all product and service-related information to be combined in a central database.
Efficiency metrics as a starting point for developing a Data Center Business Value Dashboard
Many data center managers are looking at their data center mainly from an efficiency perspective. This is an important and valid approach to identify the right data centers that support the overall strategy of a data center but is reflects only a part of the overall data center management approach that should always follow the core strategy of the business model a data center has to support. With regards to this, looking at the data center efficiency only is like trying to save fuel while driving a car but not knowing where the destination is, when I have to be there, and which different ways I can take to get to my destination. To align the data center with an underlying business strategy all parameters have to be transparent and analyzed to identify the metrics to control and manage a data center. Once the metrics are found, a Data Center Business Value Dashboard can help in bringing all the metrics together.
The Green Grid defines PUE as total facility energy divided by total IT equipment energy.
Power Usage Effectiveness, or PUE, was introduced to provide a technical measurement of the energy efficiency rate of the data center which, from a business perspective, could also be perceived as the production process. The metric can be used to compare data center designs and evaluate the impact of any change to the overall power efficiency.
Today, PUE is the basis for Green IT ratings. It is used to compare data center designs, serve as a measurement for sustainability efforts, and assure compliance with (local) legal requirements such as E-Waste regulations. Within a production process control optimization effort, PUE is the main performance indicator measuring success or failure. Values are generally determined separately for individual facilities.
Apart from determining a PUE value for a facility, a partial PUE (pPUE) is often calculated for containers or subsystems that do not take into account all of the infrastructure components in order to highlight one particular portion of a data center, such as “Container only” PUE, “Cooling system” PUE, or “Power delivery” PUE.
The concept and definition of PUE can be considered mature and well-established. Introduced by the Green Grid in 2007, the metric was eagerly and widely adopted by power equipment manufacturers who created control software that provide users with pPUE values for their subsystems. Memorandums , ,  provide explicit directives on proper assessment to produce comparable results.
One problem remains: As a purely technical metric, PUE lacks significance from a business perspective. A high or “bad“ PUE value shows that there is a problem, but gives no indication where it comes from or how to get rid of it. On the other hand, a low, respectively “good” PUE value shows that energy is used efficiently within the production process, but is not a reliable indicator that all this energy is spent wisely.
Rising energy cost and environmental impact of data center operations require on-going efforts to minimize energy waste and to spend a larger portion of energy on the data center’s end purpose, namely compute capacity, rather than on supporting systems such as cooling. While considerable technical progress has been achieved in recent years and PUE/DCiE proved useful to measure these improvements, other changes in the data center, such as the move towards virtualization, are not reflected in these indicators.
To identify optimization potential for servers, storage, and network equipment, Gartner has recently introduced metrics that take compute capacity into account and differentiate between the energy used when all systems are idle (c.f. IE, or Idle Energy) and energy consumption when all systems are operating under load (CE, or Computational Energy). The concept also takes into account that providing reliable power is essential to running a data center and some losses are inevitable in this process.
To allow your DCIM software to calculate these values automatically on the fly, additional power meters need to be connected and the analysis should include a reasonable observation period. Constantly monitoring the total power usage of your data center by an additional meter at the very top of the power chain (e.g. power meter at transformer from grid or your electrical bill meter) will allow the compute engine to reliably identify idle phases from historical data and use measured values from these periods as the baseline for further evaluation.
Together with device level power usage measurements, these metrics provide powerful and detailed insight into the efficiency of your setup; they deliver valuable information for management decisions on data center layout and hardware choice.
As these new metrics also focus on the production side and the energy efficiency at a specific site, they do not fully take into account the business value of the compute power or other business relevant components for the overall efficiency of all your data center operations.
The Data Center Business Value Dashboard
There is an ongoing trend in data centers to identify metrics and key indicator frameworks that reflect not only the technical efficiency of a data center but also the business impact and strategic orientation. The aim is to align business goals and services provided to customers with the underlying resources, assets and infrastructure. Using these frameworks, data center managers will have a more valid basis to make right business decisions and to track the data centers business performance towards the defined strategy.
There might be a thousand different potential indicators for a data center, which in turn dilute the focus. To avoid this it makes sense to transfer existing business concepts, for example the „Balanced Scorecard“ (BSC) concept, to the data center environment. This leads to the Data Center Business Value Dashboard.
The balanced scorecard is a strategy performance management tool — a semi-standard structured report supported by design methods and automation tools that can be used by managers to keep track of the execution of activities by the staff within their control and to monitor the consequences arising from these actions.
The typical characteristics that define a balanced scorecard are:
- Its focus on the strategic agenda of the organization concerned,
- the selection of a small but strategically relevant number of KPIs to monitor, and
- a mix of financial (so called lagging indicators) and non-financial (or leading indicators) items.
In its original form the balanced scorecard used a „4 perspective“ approach, namely “Financial” (financial performance), “Customer/Stakeholder” (satisfaction), “Internal business processes” (efficiency) and “Learning and growth” (knowledge, innovation and resources).
In the design phase of a balanced scorecard it is important to identify and map the relevance of strategic objectives that are important to reach an overall strategy. Typical key components of a balanced scorecard are strategic objectives, strategic linkage model and perspectives, measures, resulting key performance indicators and initiatives.
The outcome of a balanced scorecard design process is a strategic cause-and-effect chain containing all strategic objectives and their influence on each other as well as a dashboard containing between 20 to 25 relevant indicators to monitor.
Using these elements, a balanced scorecard helps managers to monitor organization performance against strategic goals.
The indicators and the strategic objectives of a balanced scorecard are always individually adapted to the strategy and vision of the organization. Some of the indicators used are very common. It is perfectly fine to change and adjust perspectives and to adapt items to the basic business model of an organization.
With this in mind, the balanced scorecard approach could be used to develop a dedicated Data Center Business Value Dashboard which can then be used to measure and track the overall performance of a data center service provider towards its strategy and vision. Such a dashboard will most likely not be re-usable for every other data center without adaptions to the needs and strategy of the data center while some of the key performance indicators used within the Data Center Business Value Dashboard might fit for many use cases.
It is also very important to keep track of the whole value chain of a data center instead of only looking at one part or subsystem — like (partial) energy efficiency — without reflecting on the backup strategy or only looking at cost performance without keeping physical infrastructure requirements in mind.
As an example, the following section shows a Data Center Business Value Dashboard based on these ideas that was developed containing indicators to follow the strategy of a banking customer’s data center described as “Industrialize IT Infrastructure Services for Banking – Effectiveness – Profitability – Efficiency”.
Starting from this overall strategy, four data center-related perspectives were developed for this example case:
- Finance: Describing the financial performance goals of the data center related to the overall data center strategy. These goals can be to increase the financial performance and to reach growth but also to reduce the general infrastructure cost that is not directly related to the value chain or service delivery.
- Markets and Products: Contains all strategic objectives that will have an important impact on the financial performance, such as a standardized product portfolio of the data center or increasing the flexibility of how services are provisioned to the customers. This perspective can also include the customer-orientation such as customer satisfaction and related influencing factors.
- Processes: Based on the financial objectives and the market and customer view, further strategic objectives can be derived which have an impact on the product standardization, agility or customer satisfaction. This perspective mostly contains important efficiency goals which have an impact on cost but also on the service provisioning times. Especially for data centers, there is huge potential to improve the operational processes, like planning and forecasting or install-, move-, add-, and change processes. If provisioning times and service delivery times can be reduced, then customers will be more satisfied. This objective can be reached through a high standardization degree and modularity in the service offering. Furthermore, it will lead to a well-defined service catalog. But also increasing or keeping a high reliability is an important objective for enterprise data centers in the financial sector.
- Data Center Infrastructure: Another important perspective for a data center is the view into the production assets which is the data center infrastructure and all related information regarding IT, network, power, space, climate and much more. All available data center assets must be planned, installed, monitored, analyzed and documented in a way to keep the lights on and to deliver the right services to the customer. But all this in an energy-efficient and secure way that is aligned with the business strategy and all strategic objectives of the data center. Increasing efficiency of the infrastructure and securing capacities are objectives often mentioned by data center service providers in the finance sector.
The example above of the strategy, perspectives and influencing strategic objectives used by the balanced scorecard methodology shows where the magic can happen for data center managers. This approach can help find the right cause-and-effect chain and the related key performance indicators to manage and control a data center not only from an infrastructure optimization level but also in an overall strategic manner that combines infrastructure and production optimization with the business and product delivery scope of a business strategy oriented and customer-focused data center over the whole value chain.
This method gives a framework that aligns the business requirements with the underlying data center infrastructure.
In the next step it is important to identify and define the relevant KPIs that will help measure, analyze and control the strategic objectives. Based on the example above, FNT has developed 27 key performance indicators that can help data center managers evaluate and discuss the accomplishment of their strategic goals.
Let’s take a closer look at several exemplary potential KPIs from each perspective of the sample scorecard that could be part of a Data Center Business Value Dashboard for data center managers, in addition to standard metrics already on your shortlist.
Cost per Application (Finance perspective)
Accrued monthly cost for running a specific application or software product.
This metric supports the scorecard perspective “Finance”. The evaluation should include all costs, such as operating system license, application license(s), maintenance, services, etc.
All necessary data to calculate the KPI can be provided through the integrated data model of FNT Command. FNT Command is a standard software that provides full transparency into all IT and telecommunications structures for managing IT assets, cabling and infrastructure, data centers, and telecommunications resources. All relevant data are available within one central database and can be easily accessed through the open architecture and connectivity layer of FNT Command.
To compute the cost, access to additional data from service management software will be necessary. FNT also offers ServicePlanet, a database that is the control center for product portfolio and service management, providing the foundation for defining, managing, and monitoring all business services and service assets in the data center or IT infrastructure over their entire service lifecycle. Standardizing products and their components in FNT ServicePlanet allows all product and service-related information to be combined in a central service management database.
You can either evaluate your data centers individually for internal comparison or calculate a single value for the service provider entity.
SLA Compliance (Market and product perspective)
Rate of compliance with existing service level agreements, expressed as a percentage.
The value is calculated from the total number of SLAs within a certain period of time used for the analysis that are fully compliant and the number of SLAs where incidents were recorded that are considered not in compliance with requirements. The gravity of issues may be taken into account and the non-compliance weighed accordingly.
This metric supports the scorecard strategy “Market and Products” and indicates whether data center operations guarantee SLA conditions and/or whether the SLA definitions associated with a service should be adjusted.
To compute this metric, FNT Command needs additional data from your service management tool (e.g. FNT ServicePlanet) and your ticketing tool. You can either evaluate your data centers individually for internal comparison or calculate a single value for the service provider entity.
Average Provisioning Time (Process perspective)
Time required provisioning a specific service or all services.
The amount of time required between a customer ordering a service (e.g. in your self-service web portal) and the instant that the service is made available to the customer (e.g. handover of OS root password). The metric can either be calculated individually for every service offering or as an average value for all services. The value might be showing net (working hours) or gross (total time span including off-hours, weekends, etc.) time spans.
This metric supports scorecard strategy “Processes”.
In general, FNT Command is used to support provisioning processes because in Command all assets and configuration items will be defined and planned. The asset database of FNT Command contains all relevant different assets that are necessary to provide a final service to the end-customer. This is the reason why FNT Command provides all data to other workflow or automation systems that is needed to automate the provisioning process itself. And also the point in time when a service is ready for use will be recorded in FNT Command’s database. To compute this metric, FNT Command might need additional data from external systems (if you are using an external workflow or portal).
You can either evaluate your data centers individually for comparison or calculate a single value for the service provider entity.
Hardware Standardization Rate (DC Infrastructure perspective)
The rate of standardization within a type of equipment such as servers or switches.
As standardization is considered essential to efficient contemporary data center operations and the reduction of the number of variants generally increases operational performance while considerably cutting costs, this KPI is an index of how close you have come to the ideal of 100%. This metric supports the scorecard perspective “DC Infrastructure”.
In a simplified model that assumes a single type of device is desirable, the rate could be defined as
HSR simplified = (CIMAX * 100) / CITOT
HSR : Standardization Rate (Simplified Model) in Percent [%]
CITOT : Total Number of all Devices of specific Device Type
CIMAX : Greatest Number of Devices of specific make and model
You run a total of 2,518 servers in your data center, 410 of which are of the exact same make and model from a purchase frame contract for standardized hardware while the remaining 2,108 devices are of 49 different types including legacy devices.
In this case, your standardization rate would be:
The metric can easily be calculated from the asset data coming from FNT Command and does not require any additional data sources. FNT Command provides extensive reporting functions for this purpose though its integrated Query Editor module. You can either evaluate your data centers individually or calculate a single value for the entire enterprise over all data center sites.
In a complex model that takes into account a number of different server models or variants that are required to provide services instead of a single type, the rate could alternatively be defined as
Many more examples of relevant KPIs can be provided by FNT, but in general they have to be analyzed regarding the goals and strategy of each data center.
Once you have identified all the KPIs that are relevant to reach your goals and to support your long-term management strategy, you design your individual balanced scorecard and add all relevant KPIs. Using our previous scorecard strategy example, this would produce the following:
The standard software products, FNT Command and FNT ServicePlanet, form the basis of FNT‘s holistic management system for services, configuration management, and infrastructure management. All important service and infrastructure information for building infrastructure, networks, physical IT assets, and applications and services is planned, documented, and managed in FNT Command using the integrated FNT data model. In many cases, data and information relating to complex systems, such as IT environments, telecommunications networks, and data centers, is not available centrally, but rather held in a number of different systems designed to handle specific tasks.
Important information about TC systems, for example, may be located in special, vendor-specific TC software products. Other data for configuring virtual environments is typically contained in specialist tools developed for this purpose, and specific data center-related information is stored in the associated building management systems or in vendor-specific monitoring solutions.
This is where FNT‘s open system architecture comes into play. By being able to exchange data with ease, it becomes possible to compare planned values and planned scenarios with real, partly measured and then imported data in the integrated FNT data model. You thus have a comprehensive overview of your valuable assets at all times and can make decisions with confidence based on valid data.
With its integrated data model and the open system architecture, FNT can provide a large amount of relevant data related to your data center infrastructure based on FNT Command but also related to your products, services, service levels, costs or pricing structure based on FNT ServicePlanet.
The integrated data model from FNT combines all these areas based on the large data model that covers the following 7 layers:
1. Facility Layer
2. Physical Layer
3. Logical Layer
4. Virtualization Layer
5. Applications Layer
6. Services Layer
7. Business Services Layer
With this, FNT provides central management and optimization software for IT, interconnectivity and data centers. From the building infrastructure (power, cooling, floor space, etc.) and IT infrastructure (such as networks, servers, and storage) down to the services (software, applications, and services), FNT software enables a comprehensive and integrated view of all valuable data center resources. From recording to monitoring using live power consumption and temperature values through to planning your entire data center, every relevant piece of information will be accessible and can be evaluated based on the integrated FNT data model.
FNT Command features powerful reporting functions to provide you with meaningful metrics, which can be made available to users through dedicated dashboards or included in enterprise business information tools. To visualize the data center strategy, the selection of KPIs can be populated from FNT Command into a Business Value Dashboard or add widgets to an existing dashboard application.
To tackle the challenges of guaranteeing reliable operations of your data center, improving efficiency while cutting costs all at the same time, strategic actions are needed. Monitoring the effectiveness of your efforts to reach your goals requires more data than today’s generally adopted technical facility metrics. Business related key performance indicators encompassing the whole value chain of your data center production should be defined to allow you to measure your success. As strategies and business models differ, the supporting KPIs need to be selected accordingly and individually.
As a leading vendor of integrated software solutions for Infrastructure Management and Service Management in data centers, IT and telecommunication, FNT Software is providing a sophisticated DCIM solution to manage all aspects of data center infrastructure and related processes. FNT’s solution also includes the option to manage cable and network infrastructures as well as logical networks and communication services. Furthermore, as all information is managed in a central repository featuring an integrated data model, which comprises all relevant asset types, this powerful solution links IT-, software- and application-related data to your physical assets.
Utilizing this unique and complete data model, FNT is the only software vendor on the market that can provide data, functional capabilities, and a proven methodology that allows service providers like data center operators to modularize, standardize, and define their product & service portfolio and align products with the underlying infrastructure.
Based on the strength of its data model combined with an open connectivity layer for unrestricted data exchange, FNT is the ideal source for key information needed in a Data Center Business Value Dashboard that can be tailored to the needs of your data center managers.