Sustainability Pillar

The Sustainability pillar was added to the Well-Architected Framework in December 2021, reflecting the growing importance of environmental responsibility in technology. It focuses on minimising the environmental impact of running cloud workloads by reducing energy consumption, improving efficiency, and making informed choices about resource usage.

---

Why Sustainability Matters in Cloud Architecture

Every computational workload has an environmental footprint. Servers consume electricity, which generates carbon emissions. Cooling systems, networking equipment, and storage devices all add to this footprint. While cloud computing is generally more energy-efficient than traditional data centres (due to economies of scale and higher utilisation rates), there is still significant room for improvement.

AWS has committed to powering its operations with 100% renewable energy by 2025 and reaching net-zero carbon by 2040 (The Climate Pledge). As a cloud architect, your design decisions directly influence how much energy your workload consumes.

---

Design Principles

The Sustainability pillar is guided by six design principles:

1. Understand Your Impact

Measure the environmental impact of your cloud workloads. Use tools like the AWS Customer Carbon Footprint Tool to track your carbon emissions. You cannot improve what you do not measure.

2. Establish Sustainability Goals

Set specific, measurable sustainability goals for each workload. For example, reduce compute hours by 20% over the next year, or increase the percentage of workloads running on Graviton processors (which are more energy-efficient per unit of compute).

3. Maximise Utilisation

Right-size your resources to avoid idle capacity. An underutilised EC2 instance wastes energy. Use Auto Scaling, serverless architectures, and managed services to ensure that provisioned resources are actively delivering value.

4. Anticipate and Adopt New, More Efficient Offerings

AWS continually introduces more efficient hardware and services. Graviton-based instances, for example, offer better price-performance and use less energy per unit of compute. Stay informed and migrate to more efficient options when they become available.

5. Use Managed Services

Managed services allow AWS to optimise resource usage across multiple customers, achieving higher utilisation rates than you could on your own. Using services like Fargate, Lambda, RDS, and DynamoDB means AWS handles the efficiency of the underlying infrastructure.

6. Reduce the Downstream Impact of Your Cloud Workloads

Consider the end-to-end impact of your workload, including the energy consumed by end-user devices. Minimise the amount of data transferred, reduce page sizes, and use efficient data formats to lower the energy cost of delivering your service to users.

---

Best Practice Areas

Region Selection

The carbon intensity of electricity varies by region. When compliance and latency requirements allow, choose AWS Regions that are powered by a higher proportion of renewable energy:

• Regions with renewable energy commitments — AWS publishes information about its renewable energy projects by region. Consider this when choosing where to deploy
• Proximity to users — while sustainability is important, deploying far from your users increases latency and data transfer, which also has an environmental cost. Balance sustainability with user experience

User Behaviour Patterns

Understanding how users interact with your application helps you optimise resource usage:

• Sustainability-aware scaling — align scaling policies with actual demand patterns. Reduce capacity during low-traffic periods
• Stop unused environments — development and test environments often run 24/7 but are only used during business hours. Schedule them to stop outside of working hours