Redfish and PLCs: Data Center Cooling and Energy Management
About Data Centers
Management technology for data centers is dynamic—and PLCs are taking part in the vital tasks of cooling and energy management, and adapting to changing trends in their technology. One growing trend is reliance on Redfish, a highly scalable RESTful API protocol that is replacing IPMI as the de facto standard for managing and monitoring computer hardware
The explosion of cloud computing, the Internet of Things (IoT), the demands of AI, and large language models such as ChatGPT, are creating great demand for data storage and processing capabilities.
The computing effort used to run the data center generates heat; excessive heat must be removed to avoid damaging servers—and this relies on cooling equipment that also runs on electricity.
Energy Trends
A 2024 report by Goldman Sachs Research forecasts that US data centers will use 8% of US power by 2030, compared with 3% in 2022, as well doubling their carbon dioxide emissions. They also note that data center workloads nearly tripled between 2015 and 2019—while holding power usage fairly stable, at about 200 terawatt-hours per year.
This was partly due to data centers growing more efficient in power use—in other words, continuing advances in technology can potentially be controlled by PLCs.
Data centers are also incorporating renewable energy sources such as solar and wind power.
Cooling Technologies
Data centers use a range of devices—such as servers, storage drives, network devices—all of which are powered by electricity and therefore generates heat that can cause equipment failure and reduce performance.
Effective cooling is essential, and so older, less efficient technologies such as computer room air conditioners and evaporative cooling are being phased out, replaced by methods such as Direct Liquid Cooling (DLC), where liquid coolant is circulated directly through server racks. Another growing trend is Liquid Immersion Cooling, where servers are submerged in a dielectric liquid, to efficiently cool the hardware. Now, AI is also integrated into data center management systems; algorithms analyze sensor data to dynamically adjust cooling processes. This ensures that energy is used efficiently while maintaining ideal operating conditions for servers. This reduces both energy costs and the carbon footprint of data centers.
The Role of Control: PLCs and the Redfish Standard
Redfish is a RESTful interface API, an interface that two computer systems use to exchange information securely over a network. Based on a JSON data format, secure, scalable Redfish enables out-of-band management of any device that has a power source and access to a network connection. Designed to manage and interact with data center hardware like servers, storage, networking equipment, and power distribution units (PDUs), Redfish enables automatic monitoring of hardware components including sensors, power supplies, fans, and chassis, and metrics such as voltage or temperature values.
Via Redfish, users can automate maintenance tasks such as setting power thresholds, perform remote recovery procedures, collect system and event logs, respond to warnings and critical events, and more.
Typically used to automate and control industrial processes, Programmable Logic Controllers (PLCs) can communicate with data centers via Redfish by leveraging the Redfish API to interact with the IT equipment and infrastructure within a data center.
Here are a few examples of how integrating Redfish with PLCs can benefit data center management.
Energy Management
- Dynamic Power Management Procedures: PLCs can automate various tasks based on the data retrieved via Redfish. PLCs can dynamically manage power loads, ensuring that energy is allocated efficiently. For instance, PLCs can balance power distribution based on the current load on servers, turning off non-essential equipment during low-demand periods.
- Fault Detection & Response: If Redfish data indicates a fault in the power system, PLCs can rapidly isolate the issue and redirect power to ensure uninterrupted operations. Redfish provides the necessary data to inform these decisions, such as identifying which components are affected and how to best reroute power.
- Remote, automatic control, based on demand: Redfish enables control commands to be sent to hardware components; a PLC can therefore use Redfish to power on/off a server or reset a device remotely. A PLC control program can automate routines, such as bringing systems online or offline based on demand.
Cooling Control
- Real-time Control: PLCs excel at managing the immediate, real-time operations of cooling systems, reliably controlling physical hardware such as adjusting fan speeds, controlling cooling units, or managing liquid cooling pumps , ensuring that cooling is precisely matched to the current thermal load.
- Sensor Data – Redfish-PLC integration: Redfish collects data from environmental sensors across the data center, such as temperature, humidity, and airflow rates; PLCs can access this sensor data via the Redfish API to determine the current thermal conditions within the data center.
- Remote Management: Redfish Redfish Data center cooling enables remote control of cooling equipment, which can be managed by the PLCs for real-time adjustments. This can include commands to adjust airflow, activate cooling zones, or change the power state of cooling equipment.
- Environmental control: in a data center, PLCs can be integrated with the Redfish API to monitor and control environmental conditions, such as temperature, humidity, and power distribution.
- Actions and alarms: via the Redfish API, PLCs can harvest real-time data from data center hardware components—and use it to trigger specific actions based on predefined conditions, for example adjusting cooling systems or raising alarms if the temperature exceeds a specific threshold.
Data Collection and Analysis
- Redfish provides a wealth of information about data center hardware. PLCs can collect this data over time and analyze trends, which can be used for predictive maintenance, optimizing power usage, and improving overall efficiency.
- The PLC can also log data for compliance and auditing purposes, ensuring that all actions taken within the data center are recorded and traceable.
Summary
Via Redfish, PLCs can further benefit systems by interfacing with other data center management systems (such as DCIM—Data Center Infrastructure Management) to provide a comprehensive view of the data center’s operations. Integrating with other systems enables PLCs to contribute to a holistic approach to data center management, ensuring that all aspects of the environment are optimized and running efficiently.
In summary, PLCs can interact with data centers via Redfish by leveraging the API to monitor, data center operations, ensuring efficient and reliable management of the infrastructure. This integration allows for enhanced automation, real-time monitoring, and intelligent decision-making in data center environments.