View a markdown version of this page

DRHCSUS03-BP01 Monitor workload and component resource utilization to identify any that are unneeded or over-provisioned when using Local Zones - Data Residency and Hybrid Cloud Lens

DRHCSUS03-BP01 Monitor workload and component resource utilization to identify any that are unneeded or over-provisioned when using Local Zones

Monitoring workload utilization is critical to identifying those which are overprovisioned and can be scaled down to reduce energy consumption.

Desired outcome: Workloads are deployed on Amazon EC2 instances that are aligned to requirements and provide optimal performance.

Benefits of establishing this best practice: Workloads and resources which are overprovisioned can be scaled down to reduce energy consumption and support sustainability objectives.

Level of risk exposed if this best practice is not established: Medium

Implementation guidance

Monitor and capture workload metrics such as CPU utilization, memory utilization, network connections, and network throughput using Amazon CloudWatch metrics to identify workloads or components that are no longer used or are under-utilized. Routinely review the recommendations made by AWS Compute Optimizer to identify instances that have been over-provisioned, and adjust as necessary to match instance type and size to the workload requirements. With AWS Local Zones, it is possible to maximize sustainability by using the large variety of Amazon EC2 instances to match resource consumption to workload requirements.