Skip to main contentIBM Z Hot Topics

IBM zSystems and IBM LinuxONE – Helping You Meet Your Sustainability Commitments

IBM has a decades-long history of leadership in sustainability and continues that excellence with the new IBM z16 and IBM LinuxONE 4 enterprise servers. Learn about new capabilities to support power consumption monitoring at the more granular partition level and how to consume the data either on the new HMC Environmental Dashboard or through the enhanced HMC Web Services API.

IBM zSystems and IBM LinuxONE – Helping You Meet Your Sustainability Commitments

Introduction

Have you noticed that every business these days has something to say about sustainability? Your online airline booking now tells you how much carbon dioxide equivalent (CO2e) your trip will produce. The box for your new sneakers proclaims how its materials were sustainably sourced and are fully recyclable. The website you visit to learn about a particular product doesn’t even mention the product until after you scroll past all the sustainability declarations. Businesses work hard to be more sustainable, and they want to make sure you know it.

Both consumers and businesses consider sustainability fundamental to remaining competitive. The majority of organizations see environmental sustainability as one of their highest near-term priorities, and Environmental, Social and Governance (ESG) issues have become board-level discussions for most organizations. In many cases, consumers are willing to change their purchasing decisions based on sustainability factors, and businesses are only considering buying from and partnering with other businesses that share similar goals of protecting the environment. To help harness the competitive advantage of sustainability, organizations are committing to accelerating the transition with commitments to reach net zero emissions before 2050.

Of course, a lot of the growing focus on sustainability in business is driven by corporate mandates. There are now more than 2,500 climate laws and policies worldwide that look at reducing greenhouse gas (GHG) emissions. Businesses meet the associated requirements by setting goals for procuring electricity from renewable power sources, designing products for increasing energy efficiency, ensuring that source materials are responsibly sourced, and so on. But how do businesses prove that they are meeting those goals? One common answer is by collecting data and then reporting the metrics to the person who manages adherence to the business’ sustainability requirements, oftentimes a Chief Sustainability Officer (CSO).

To help support the needs to report on the power consumption of used or offered IT services, IBM z16™ and IBM® LinuxONE 4 introduce brand-new capabilities for collecting the data at the finer granularity of a logical partition. You can then access the data where it makes sense, either on the new Hardware Management Console (HMC) Environmental Dashboard or through the HMC Web Services API.

IBM’s history of strong leadership in sustainability

In January 2022, the IBM Institute for Business Value (IBV) conducted a study to understand how organization executives are addressing sustainability. The study found that 13% of executives are serving as trailblazers among their peers, helping to ensure that environmental sustainability is front and center as an enterprise strategy and working to integrate sustainability within digital transformation. However, the study also found that even though 86% of the companies in the study have a strategy for sustainability, only 35% have taken any significant action on that strategy. A lot of businesses just don’t know where to start.

IBM has been protecting the environment for more than 50 years through the management of its operations and the development of its products. In 1971, IBM formalized its environmental programs and emphasized its commitment to environmental leadership by issuing its first corporate environmental policy, a quarter-century before the first International Organization for Standardization (ISO) 14001 environmental management systems standard was published in September 1996.

We haven’t slowed down since.

  • In 1997, IBM became the first major multinational company to earn a single global registration of its environmental management system (EMS) to the ISO 14001 standard.
  • In 2007, IBM first published its policy position on climate change.
  • In 2015, IBM voiced its support for the Paris Agreement and reaffirmed its support in 2017.
  • In 2018, IBM established a second-generation goal for the use of renewable energy and a fourth-generation goal to reduce carbon dioxide emissions. In addition, IBM expanded the scope of these goals by adding the energy use and CO2 emissions associated with data centers located in facilities managed by third parties where IBM does not procure the energy.
  • In 2019, IBM became a founding member of the Climate Leadership Council and publicly supported the Council’s plan for a carbon tax, with the proceeds of that tax serving as a “carbon dividend” to be returned to citizens.
  • In 2021, IBM committed to reaching net zero GHG emissions by 2030.
  • In 2022, IBM served as a technology partner of the 2022 United Nations Climate Change Conference (COP27).

Figure 1 lists just a few of IBM’s key environmental milestones related to the development of our products.

Figure 1

Figure 1: IBM’s history of environmental leadership

A legacy of impact for IBM zSystems and IBM LinuxONE

Being at the leading edge of sustainable product design isn’t new to IBM zSystems™ and even IBM® LinuxONE, which is the Linux-only enterprise server offering that is based on the same architecture and shares all the same hardware innovations as zSystems. As far back as IBM System z9®, we offered power monitoring and power estimation tools. With IBM z13® and IBM® LinuxONE Emperor, we started supporting the ASHRAE A3 environmental envelopes, making it easier for data centers to reduce their reliance on mechanical cooling. And with the IBM z14® Model ZR1 and IBM® LinuxONE Rockhopper II, we completely transformed the mainframe, moving to a 19” frame form factor that makes it easy to fit into modern data center cooling containment systems for hot aisle/cold aisle configurations. As part of the transformation, we also significantly improved energy efficiency, which has been a cornerstone of zSystems for the last 28 years. This means that, with every generation we released, even though the machines grew in capacity, the capacity per kilowatt became more efficient to the point that it has increased over 100x since the first generation. Others are taking notice of the sustainability aspects of zSystems and LinuxONE, too. As acknowledgement of their impacts on a more sustainable future, both brands have been honored with the industry-wide SEAL Sustainable Product Award, which is given to innovative and impactful products that are purpose-built for a sustainable future. z15 was awarded in 2021, and IBM® LinuxONE Emperor 4 won in 2022.

Energy management capabilities

Figure 2 provides a summary of the rich set of energy management capabilities that zSystems and LinuxONE have had for generations. These are becoming essential tools as businesses need to be able to report on the energy use of their infrastructure to prove that they are making progress on their net zero commitments. With pending regulation such as the US Securities and Exchange Commission’s ESCRD and the European Union’s CSRD requiring GHG emission reporting for enterprises, having the ability to easily report on the power consumption of your infrastructure is paramount.

Figure 2

Figure 2: zSystems and LinuxONE history of energy management innovation

The energy management features of zSystems and LinuxONE allow for the integration into modern data center infrastructure management (DCIM) systems using a set of secure, REST-based Web Services APIs. The HMC Web Services API metrics groups allow the system to report key environmental and power consumption data that includes ambient temperature and humidity, heat load, exhaust temperature, system power consumption, and power on each line cord phase. This data is aligned with the latest recommendations from ASHRAE – the ASHRAE Tier 1 classification.

The zhmcclient package is a client library written in pure Python that interacts with the Web Services API. Available on Github, the goal of this package is to make the HMC Web Services API easily consumable for Python programmers. See the Resources section for a link to the package. There is also a Prometheus exporter that makes it easy to set up custom metric collections.

New sustainability capabilities with z16 and LinuxONE 4

z16 and LinuxONE 4 of course continue the tradition of providing superior scalability, security, and performance for your data-serving needs. They also introduce some pretty cool new capabilities to help you meet your sustainability goals.

More granular power reporting

We’ve already said that organizations need to be able to report on the power that their infrastructure consumes. This requires a finer level of granularity in power reporting than was previously available. z16 and IBM® LinuxONE Rockhopper 4 introduce the ability to conduct power reporting by logical partition (LPAR). Because truly understanding how the power is used can be complicated, we do the work for you. This new capability breaks down the power into three categories for you:

  • Partition power: Power utilized by components (CPU, memory, I/O) that are assigned to individual partitions
  • Infrastructure power: Power utilized by infrastructure components including infrastructure switches, Support Element (SE) / Hardware Management Appliance (HMA), and power distribution units (PDU)
  • Unassigned power: Power utilized by unused CPU, memory and I/O adapters, and components that are not assigned to any partition (including standby components)

Figure 3

Figure 3: Breaking down the system power consumption

By breaking down the power into the categories shown in Figure 3, you have the ability to account for the consumption in a way that works best for your business.

You can then consume the power usage data graphically in the new HMC Environmental Dashboard experience or the HMC Monitors Dashboard, or programmatically through the HMC Web Services API.

A new environmental dashboard experience

HMC users can access system environmental data through various graphical user interfaces. To simplify this, starting with z16 and LinuxONE 4, the new HMC Environmental Dashboard experience, which replaces the former HMC Environmental Efficiency Statistics task, is available. This new experience provides you with the following capabilities:

  • Monitoring power consumption at the system level and now also the partition level
  • Selecting time ranges and metrics to view historical data
  • Charting and analyzing data
  • Exporting data for use elsewhere

The dashboard is optimized for the new ability to collect power usage at the partition level.

Want to see what this new experience looks like? See Figure 4 and then also see the Resources section for a link to the HMC Environmental Dashboard experience demo video that you can check out.

Figure 4

Figure 4: The new HMC Environmental Dashboard

Expanded HMC Web Services API

The new HMC Environmental Dashboard simplifies your access to power consumption information. However, more often, this data is needed in other reporting tools, such as DCIM solutions, observability tools like Instana, or ESG reporting software like Envizi. To support these use cases, we’ve enhanced the HMC Web Services API.

This article provides just a summary of the new capabilities. You can find full details on the API implementation in the Hardware Management Console Web Services API publication.

This implementation adds three new reporting capabilities.

  • CPC historical sustainability data
  • Logical partition historical sustainability data
  • Partition historical sustainability data

All of these return an array of data over a user-specified time range. This allows for the capture of historical data and, if instantaneous data reporting is not required, minimizes the number of API calls needed to get the sustainability data from the HMC.

Get CPC Historical Sustainability Data

The Get CPC Historical Sustainability Data operation retrieves the energy management related metrics that are listed in Table 1, all for a specific time range. In addition to the preexisting total system power metric, we added three new power metrics to provide finer granularity to the power reporting.

  • total-wattage: The total system power (previously existed)
  • infrastructure-wattage: The power consumed by the support elements, Ethernet switches, and cooling
  • partition-wattage: The sum of the power of the CPU, memory, and I/O resources used across all partitions
  • unassigned-wattage: The power consumption of all CPU, memory, and I/O resources not assigned to a partition

Format:

POST /api/cpcs/{cpc-id}/operations/get-historical-sustainability-data

In this request, the URI variable {cpc-id} is the object ID of the CPC object. On successful completion, the response body is a JSON object with the fields listed in Table 1.

Table 1: CPC Historical Sustainability Data

Field nameDescription
total-wattageEach element of this array is an integer-data-point object representing the total system power, in watts, at a specific point in time.
partition-wattageEach element of this array is an integer-data-point object representing the power consumption sum of all partitions, in watts, at a specific point in time.
infrastructure-wattageEach element of this array is an integer-data-point object representing the total infrastructure power (power consumed by support element, Ethernet switches, and cooling) in watts, at a specific point in time.
unassigned-wattageEach element of this array is an integer-data-point object representing the total power consumption of resources not assigned to a partition, in watts, at a specific point in time.
heat-loadEach element of this array is an integer-data-point object representing the amount of heat in Btu/h removed from the system per hour, at a specific point in time.
heat-load-forced-airEach element of this array is an integer-data-point object representing the amount of heat per hour removed from the system by forced-air, at a specific point in time.
processor-utilizationEach element of this array is an integer-data-point object representing processor utilization in percentage, at a specific point in time.
ambient-temperatureEach element of this array is a float-data-point object representing temperature in degrees Celsius as measured by the system, at a specific point in time.
exhaust-heat-temperatureEach element of this array is a float-data-point object representing exhaust air temperature in degrees Celsius, at a specific point in time.
dew-pointEach element of this array is an integer-data-point object representing dew point in degrees Celsius, at a specific point in time.
ambient-humidityEach element of this array is an integer-data-point object representing the amount of water vapor in percentage, at a specific point in time.

Get Logical Partition Historical Sustainability Data

The Get Logical Partition Historical Sustainability Data operation retrieves the logical partition data values that are listed in Table 2, all for a specific time range.

Format:

POST /api/logical-partitions/{logical-partition-id}/operations/get-historical-sustainability-data

In this request, the URI variable {logical-partition-id} is the object ID of the logical partition object. On successful completion, the response body is a JSON object with the fields listed in Table 2.

Table 2: Logical Partition Historical Sustainability Data

Field nameDescription
wattageEach element of this array is an integer-data-point object representing the estimated power consumed by the partition, in watts, at a specific point in time.
processor-utilizationEach element of this array is an integer-data-point object representing partition processor utilization in percentage, at a specific point in time.

Get Partition Historical Sustainability Data

The Get Partition Historical Sustainability Data operation retrieves the partition data values for DPM-enabled systems that are listed in Table 3, all for a specific time range.

Format:

POST /api/partitions/{partition-id}/operations/get-historical-sustainability-data

In this request, the URI variable {partition-id} is the object ID of the partition object. On successful completion, the response body is a JSON object with the fields listed in Table 3.

Table 3: Partition Historical Sustainability Data

Field nameTypeDescription
wattageArray of integer-data-point objectsEach element of this array is an integer-data-point object representing the estimated power consumed by the partition, in watts, at a specific point in time.
processor-utilizationArray of integer-data-point objectsEach element of this array is an integer-data-point object representing partition processor utilization in percentage, at a specific point in time.

Partition power calculation

Let’s take a closer look at how the partition power share is calculated. The calculation methodology is based on allocating the measured total system power across the three categories described earlier: partition power, infrastructure power, or unused resource power. Figure 5 provides a general overview of the methodology.

For z16 and LinuxONE 4, additional hardware instrumentation was added so that the allocation of the power could be based on real time measurements of various hardware subsystems. For example, within the central processor complex (CPC) drawer, measurements are done at the levels of both the drawer through the power supply units as well as individual components inside the drawer, such as the memory. This data is gathered by the Support Element and combined with machine and partition configuration data, such as how many CPs are assigned to each partition. Because resources can be shared across LPARs, the SE power management utility then uses this data to calculate the partition power share based on its configuration and utilization of resources.

For resources that are shared, such as the processors, the reported power allotted to the processor resources takes into account the processor-percent-usage for all processors in the partition. This is used to create a utilization factor that is multiplied by the weighted processor share. For I/O cards, where we do not know exactly the share of an I/O card in use by a particular partition, we assume that shared I/O cards are shared equally among partitions that are assigned to that card. Furthermore, we can’t individually measure the power of each individual I/O adapter. But we can measure the power of the PCIe+ I/O drawer. We combine this with knowledge of the power of each I/O card type based on lab measurements (the same data used in our Power, Weight, and Airflow estimation tool). This allows us to calculate a weighted share of the card and also take into account the power characteristics of each card type.

Using this methodology allows us to calculate the share of resources that are used by each LPAR on the system. Calculating the unassigned power and the infrastructure power are straightforward after the total partition power is known. Finally, to validate that our calculation methodology adds up to the total system power, we conduct internal validation checks based on additional power measurement using the power distribution units. These include the power of the Support Elements, Ethernet Switches, and Radiator Cooling Unit (if applicable).

This methodology for distributing the system power across the LPARs is a big first step toward providing you and your customers with the level of data needed for your reporting.

Figure 5

Figure 5: LPAR power allocation methodology

Summary

The day when companies must focus on sustainability as an important part of their business has arrived. Not only being sustainable but also proving sustainability is critical, and the ability to track and report on the sustainability aspects of your data centers is a key part of that second goal. To help support your infrastructure power consumption reporting needs, the brand-new capabilities introduced with z16 and LinuxONE 4 provide the ability to collect and report power consumption with much more granularity – at the level of a logical partition. You can then access the data where it makes sense, either with the new Hardware Management Console Environmental Dashboard or through the HMC Web Services API. This new reporting capability greatly simplifies the task of breaking down the total power of a zSystems or LinuxONE enterprise system and is done “under the covers” so you can easily provide the required level of reporting and then go back to focusing on operating your business.

Resources

About the authors

Dustin Demetriou is a Senior Technical Staff Member and leads sustainability and data center innovation for IBM Systems. He focuses on product sustainability and lifecycle assessment, data center infrastructure management, and future trends and regulations. He holds a Ph.D. in Mechanical and Aerospace Engineering from Syracuse University with a focus on thermal management and energy conversion systems. He is a globally recognized expert in the field of thermal management and data center energy efficiency and has taught courses on these subject around the world.  

Susan Shumway is the Sustainability Product Manager for IBM zSystems and IBM LinuxONE enterprise servers, bringing visibility to sustainability and driving innovation to achieve market leadership for her products. After earning a BS in Engineering and an MS in Technical Communication from Rensselaer Polytechnic Institute, she joined IBM and has spent over two decades in various rewarding roles across development, project management, and product management.

Thomas Pohl is a Senior Technical Staff Member in the IBM zSystems firmware development team. He joined IBM in 2007 and has been driving the IBM zSystems sustainability development roadmap through multiple system generations. In addition, he leads efforts to reinvent and modernize IBM zSystems firmware structures. He also gives lectures, acts as a University Ambassador, and is a member of the IBM Technical Expert Council.

Sergio Soto Ortega leads, as a complex function owner, development activities for customer critical sustainability and energy management functionality for IBM mainframes. He also serves the role of test strategy focal point, driving test automation transformation in IBM Z firmware. He joined IBM in 2016 and holds a BS in Electronic Systems Engineering, an MS in Intelligent Systems, and a second MS in Computer and Communication Technologies.