Select the best compute options for your workload

Selecting the most appropriate compute option for your workload allows you to improve performance, reduce unnecessary infrastructure costs, and lower the operational efforts required to maintain your workload. The optimal choice hinges on various factors, including application design, usage patterns, and assigned configurations.

Best Practices

Leverage Auto Scaling

• Implement Auto Scaling to adjust compute resources based on demand. This helps to ensure that you are using the optimal amount of resources without over-provisioning.
• Auto Scaling can improve performance by making sure that your application can handle peak loads efficiently and scale down during low usage to save costs.
• Set up scaling policies based on metrics that reflect the real-time load on your application, such as CPU utilization or request latency.

Choose the Right Instance Types

• Evaluate different EC2 instance types to identify the best fit based on your workload characteristics, such as compute, memory, and storage needs.
• Utilize AWS Compute Optimizer to receive recommendations on the optimal instance types based on historical usage patterns.
• Consider using Spot Instances for non-critical workloads to reduce costs while maximizing performance.

Utilize Managed Services

• Where possible, utilize managed services such as AWS Lambda, Amazon ECS, or AWS Fargate to abstract away infrastructure management and improve performance efficiency.
• These services automatically scale to meet demand and optimize provisioning, allowing you to focus on your application logic rather than the underlying infrastructure.
• Evaluate the trade-offs of using managed services versus traditional EC2 instances to determine the best approach for your workload.

Optimize Resource Configuration

• Regularly review and optimize the configuration of your compute resources, including CPU, memory, and networking settings, to ensure they are aligned with performance requirements.
• Use performance monitoring tools to gather insights on resource usage and make data-driven decisions about scaling and configurations.
• Perform load testing to understand how your application performs under different configurations and adjust accordingly.

Consider Serverless Architectures

• Evaluate if transitioning parts of your architecture to serverless technologies, such as AWS Lambda, can enhance performance and reduce management overhead.
• Serverless options automatically handle scaling and resource allocation, allowing you to focus on code and functionality rather than the infrastructure.
• Assess the latency and performance implications of a serverless architecture for your specific workload.

Questions to ask your team

• What are the specific compute requirements of your application and how have you determined them?
• Do you regularly evaluate the performance of your existing compute resources against your workload demands?
• Have you considered auto-scaling options to handle variable workload needs?
• Are you utilizing multiple compute options (e.g., EC2, Lambda, ECS) efficiently within your architecture based on the application requirements?
• How do you monitor and analyze the performance of your compute resources?
• What processes are in place to identify and migrate to more suitable compute resources as your workload evolves?

Who should be doing this?

Cloud Architect

• Assess workload requirements and analyze application design.
• Evaluate different compute resources (e.g., EC2, Lambda, ECS) based on performance metrics.
• Design architectures that integrate various compute options for optimal performance.
• Monitor usage patterns and adjust compute resources as necessary to maintain efficiency.

DevOps Engineer

• Implement selected compute options in the deployment pipeline.
• Automate scaling and resource allocation to match workload demand.
• Configure monitoring tools to track performance and resource utilization.
• Collaborate with the Cloud Architect to ensure alignment with design principles.

Cost Management Analyst

• Analyze cost implications of selected compute resources.
• Recommend adjustments to reduce costs while maintaining performance.
• Provide insights on optimizing infrastructure spend through efficient compute choices.
• Regularly review spending patterns and give feedback for future resource selections.

Application Developer

• Design and build applications with performance efficiency in mind.
• Collaborate with the Cloud Architect to understand the impact of compute choices on application performance.
• Leverage specific features of selected compute options to enhance application behavior.
• Test application performance across different compute resources and optimize accordingly.

What evidence shows this is happening in your organization?

• Compute Resource Selection Checklist: A detailed checklist for evaluating and selecting the appropriate compute resources for various workloads based on application design, usage patterns, and scaling requirements.
• Performance Optimization Report Template: A template for documenting and analyzing the performance of compute resources in a workload, including key metrics and recommendations for optimization.
• Compute Resources Strategy Guide: A strategic guide outlining best practices for selecting and using compute resources, including considerations for cost, performance, and operational efficiency.
• Resource Selection Diagram: A visual diagram illustrating various compute options (e.g., EC2, Lambda, ECS) and how to choose between them based on workload requirements and architectural designs.
• Performance Efficiency Playbook: A playbook that provides step-by-step approaches for assessing compute options, conducting performance tests, and iterating on compute resource selection to maximize efficiency.

Cloud Services

AWS

• EC2 Auto Scaling: Automatically adjusts the number of EC2 instances in response to demand, ensuring optimal performance and efficient use of resources.
• AWS Lambda: A serverless compute service that automatically scales in response to events, allowing for efficient handling of varying loads without provisioning underlying infrastructure.
• AWS Compute Optimizer: Provides recommendations for optimal AWS compute resources based on your usage patterns, helping you select the right instance types and sizes.
• AWS Fargate: A serverless compute engine for containers that allows you to run containers without managing servers, optimizing resource utilization based on the workload.

Azure

• Azure Virtual Machines: Provides scalable computing resources that can be easily adjusted to meet demand, optimizing pricing and performance.
• Azure App Service: A fully managed platform for building, deploying, and scaling web apps, offering automated scaling to match application demand.
• Azure Advisor: A personalized cloud consultant that provides best practices and recommendations for optimizing Azure resources, including compute.
• Azure Functions: A serverless compute service that allows you to run code without provisioning servers, scaling automatically based on demand.

Google Cloud Platform

• Google Compute Engine: Computing resources that you can configure and scale easily, helping to balance performance and cost based on workload needs.
• Google Kubernetes Engine: A managed environment for deploying, managing, and scaling containerized applications, optimizing resource usage automatically.
• Google Cloud Functions: A serverless execution environment that enables you to run code in response to events, automatically scaling as needed.
• Google Cloud Operations Suite: Monitoring and performance tools that help analyze and optimize your compute resources, ensuring high performance and efficiency.

Question: How do you select and use compute resources in your workload?
Pillar: Performance Efficiency (Code: PERF)