Cloud optimization is performed by applying one or more of the following operations to IT infrastructure:
- Downscale
- Scale down Auto Scaling Groups’ scaling parameters to align with actual application demand patterns
- Downsize
- Migrate the workload to smaller, less expensive instance size (within the same instance family) because the current instance is not fully utilized
- Downsize – Optimal Family
- Downsize the cloud instance and at the same time move the workload to a better-suited instance family to better serve the workload. For example, move a workload from a general purpose m5.xlarge to a memory optimized r5.large
- Modernize
- Leverage newer, cheaper instance types
- Modernize – Optimal Family
- Modernize the cloud instance and at the same time move the workload to a better-suited instance family to better serve the workload. For example, move a workload from a general purpose m4.xlarge to a memory optimized r5.xlarge
- Terminate
- Turn off unused systems
- Upscale
- Scale up Auto Scaling Groups’ scaling parameters to align with actual application demand patterns
- Upsize
- Move the workload to a larger instance to avoid performance risks
- Upsize – Optimal Family
- Upsize the cloud instance and at the same time move the workload to a better-suited instance family to better serve the workload. For example, move a workload from a general purpose m4.medium to a memory optimized r5.2xlarge
Cloud Optimization Methods Unique to Densify
For all recommendations, Densify includes sizing and overall cost impact individually and for logical groups of systems. No user configuration is necessary because the Densify Densification Advisor works with the end user to determine if there are recommendations or instance types that are out of scope; analytics are tuned accordingly.
Densify’s results are based on its patented ability to match predicted workload demand with normalized cloud supply.