5 ways devops can reduce energy consumption
5 ways devops can reduce energy consumption
When we build and deploy microservices, apps, and databases, common operational concerns include their reliability, performance, scalability, and security. Perhaps it’s time we add sustainability to the list.
Business and technology executives expect IT to support the organization’s environment, social, and governance (ESG) objectives, and digital transformation programs often have sustainability goals.
In the 2022 Nash Squared Digital Leadership Report, 44% of boards recognize technology as crucial to improving their organization’s sustainability and carbon footprint. The S&P Global Market Intelligence report on the intersection between digital transformation and energy transition says addressing power efficiency and supporting ESG goals were identified as critical digital transformation drivers by more than 40% of executive respondents.
Looking at power-hungry legacy data centers and end-of-life infrastructure is an obvious first step in reducing energy consumption. Devops teams should also consider continuous improvements and finding innovative ways to address sustainability objectives. Here are five considerations.
Measure energy utilization by application
“You can’t improve what you don’t measure,” an often quoted objective attributed to Peter Drucker, lays out a good first step for IT and devops teams.
Ciaran Forde, business manager of data centers at Eaton, agrees, “The first step in energy reduction is power monitoring and measurement. Only with operational insight down to the application level can one begin understanding the who, what, where, and when of energy consumption.”
Public clouds have options for measuring energy consumption: Amazon’s free customer carbon footprint tool, Google’s carbon footprint report, and Microsoft’s sustainability calculator.
Forde shares, “One option is the native integration of energy metrics into a given application, or ensuring the collation and availability of the data for API connectivity into other specialized energy management applications.”
Upgrade power-heavy infrastructure and consider cloud options
On-premises data centers should include power consumption as an indicator to accelerate the infrastructure’s refresh or to consider architecting a modernized, sustainable solution.
Anant Adya, executive vice president at Infosys Cobalt, says, “In today’s era, it is imperative that devops teams prioritize sustainability. To do so, I strongly recommend taking the following steps: reducing resource-heavy infrastructure and moving network compute storage to the cloud, reducing the number of clouds and data centers they rely on, and moving edge locations to the cloud.”
Mike Jackson, global director of product, data center, and distributed IT at Eaton, says hybrid cloud offers sustainable options and compute flexibilities. “Now that we’re seeing the expansion between the core data center, cloud-hybrid, and edge computing, the next decision becomes selecting the optimal place to run an application.”
Devops teams are not often involved in where to locate data centers, edge infrastructure, remote offices, or factory floors, but there is room to influence these decisions. Jackson suggests asking, “Can it be run in a place where a more preferred energy source is being consumed, like renewables, or within a much more efficient infrastructure?”
Reduce underutilized cloud resources
“Just put it up on the cloud, and we can figure out how to automate consumption and configure elastic computing at another time” may be something devops engineers hear from agile product owners. The business teams might want to deploy sooner, but does that have to come at a cost to sustainability goals and devops best practices?
“Devops teams must take a lean-agile approach to effectively balance the business needs of velocity against operational sustainability demands,” says Brian Copeland, executive director at TEKsystems. “In doing so, organizations can eliminate unused and underutilized resources to help reduce the energy consumption risks associated with cloud operations.”
Some devops best practices that address sustainability goals are configuring the infrastructure as code, considering serverless computing architectures, and using elastic computing that ramps up and down infrastructure based on usage and computing demands.
Devops teams should combine elastic architecture with reporting on power consumption. “Devops teams must consider adopting cloud-native software, calculating server utilization, and monitoring power usage effectiveness,” says Adya. “With these practices, devops teams will realize the full potential of the cloud to achieve their organization’s sustainability goals and reduce energy consumption.”
Expand automation, alerts, and visualizations to save power
Once devops teams deploy apps to production, they should use reporting and tools to provide feedback on whether energy consumption aligns with original projections.
“To align with company sustainability goals, IT departments can make dramatic reductions in their electricity use by leveraging intelligent automation and resource management,” says Jeff Kukowski, CEO of CloudBolt Software. “With an advanced automated alert and visualization system, developers and other organizational stakeholders can always be informed of the environmental impact of the decisions they make throughout their day.”
Alerts and visualizations can be very useful when configuring and deploying dev and test environments where devops and data science teams are likely to overprovision computing resources. The developer doesn’t want to be slowed down by the infrastructure but should also consider cost and power consumption in sizing computing needs.
Kukowski says, “Suppose a developer is provisioning a public cloud resource and a less energy-intensive option is available. In that case, they could receive a notification alerting them to the issue and suggesting the greener option.”
Select energy-efficient data storage options
Apps and infrastructure consume power, but storing and managing big data repositories does too.
John Wheeler, senior advisor of risk and technology for AuditBoard, says, “The tendency to amass or even hoard data often results in wasteful energy consumption through needless data storage and backup.”
A data retention policy can account for data storage’s cost and energy consumption. Wheeler adds that data hoarding also increases the risk of data loss, misuse, or legal liability. He suggests, “Devops teams should collaborate with their business partners to create mechanisms to reduce data storage needs as new digital applications, products, and services are released.”
Roman Golod, CTO and cofounder of Accelario, shares one energy efficiency operation devops teams and database administrators should consider. “To meet the demands of devops teams, DBAs generally create a few copies of the production database, which significantly increases storage utilization and energy usage along with the operational costs,” he says. “By using database virtualization modules as part of the devops platform, teams can create multiple copies using their golden image and with that, significantly reduce any cost associated with storage energy consumption.”
So, whether you are an architect, devops engineer, DBA, or part of an IT operations team, you can influence and contribute to your organization’s sustainability goals.