Choose the right cloud products and resource specifications to meet your business needs. When you design application workloads and select resources, consider stability, performance, security, compliance, and operational efficiency in addition to cost. You must balance these goals against cost.
A suitable application workload design is not just about achieving a low cost. You must first fully understand all your requirements. A low-cost approach often requires compromises in other areas. After you understand all the requirements, you can prioritize them based on your business characteristics to make a comprehensive decision.
Cost and stability
If your business requires high stability, your application workload design will have higher resource costs. For example, strict requirements for metrics such as Service-Level Agreement (SLA), Recovery Time Objective (RTO), and Recovery Point Objective (RPO) increase costs. To achieve high availability, application services require more resources for cross-zone or cross-region deployments. Cross-region data replication also increases costs.
Layer your application workloads based on business importance, determine stability priorities, and select suitable stability solutions. For online applications in a production environment, choose high-stability solutions such as active-active and active geo-redundancy. For important data, use methods such as zone-redundant storage and cross-region backup to ensure data security. For staging environments, consider lowering stability requirements to reduce costs.
For more information about stability, see the Stability Pillar.
Cost and performance
Application workloads with high performance requirements often cost more. Set performance requirements for your workloads based on business characteristics. Choose suitable cloud products, resource specifications, and billing methods to balance performance and cost. The following are some suggestions for cost optimization and performance:
Run performance tests during the resource selection phase. Increase the purchase volume only after the performance requirements are met.
Implement elasticity. Monitor application workloads, resource utilization, and peak traffic periods to dynamically provision and use resources on demand.
Choose a billing method that suits your business characteristics. For example, for services that require high public bandwidth and have stable peak traffic, use the pay-by-bandwidth method and purchase Internet Shared Bandwidth.
For certain business scenarios, use burstable instances to handle performance bursts and optimize costs.
For more information about performance, see the Performance Efficiency Pillar.
Cost and security compliance
Application workloads with high security requirements often lead to high costs. Set security requirements for your application workloads based on business characteristics. Do not compromise on the security and compliance of your application workloads. The following are some suggestions for cost optimization and security compliance:
Compliance audit: Use products such as control policies, CloudConfig, and ActionTrail to easily achieve continuous compliance in the cloud. These products are currently free of charge, subject to the billing details on the official website.
Network security: Use Cloud Firewall for network protection. It supports security management across multiple accounts, so you do not need to purchase it for each account.
Key security: Use Key Management Service for centralized key management. It supports instance sharing across multiple accounts, so you do not need to purchase it for each account.
For more information about security and compliance, see the Security Pillar.
Cost and operational excellence
Building observability systems, standard change processes, and automated workflows are important parts of operational excellence. This usually increases costs initially, but the costs decrease over time. Effective cost management in the cloud is also a key part of improving operational efficiency. The following are some suggestions for cost optimization and operational excellence:
Use cloud-native Platform as a Service (PaaS) monitoring products to avoid the maintenance costs and stability issues of self-managed infrastructure.
Use infrastructure automation orchestration tools such as Terraform to build automated workflows. Turn your infrastructure into code to reduce development and O&M costs.
For more information about operations, see the Operational Excellence Pillar.