> ## Documentation Index > Fetch the complete documentation index at: https://docs.kubex.ai/llms.txt > Use this file to discover all available pages before exploring further. # Auto Scaling Group Optimization [Cloud Optimization] export const InlineImage = ({src, href, alt = "", size = "medium", customSize, position = "left", children}) => { const sizeClasses = { small: "w-6 h-6", medium: "w-8 h-8", large: "w-12 h-12", xlarge: "w-16 h-16" }; const ImageComponent = () => {alt}; const imageElement = href ? : ; return <> {position === "left" && imageElement} {position === "left" && " "} {children} {position === "right" && " "} {position === "right" && imageElement} ; }; The policy settings in this group are related to optimizing auto scaling groups. ## Minimum Scale Group Size
Description This setting defines the minimum allowed number of EC2 instances to be recommended for an ASG.
Use Case  
Values/Units An integer indicating the minimum scale group size.
Dependencies  
Example  
## - Maximum Increase
Description This setting defines the maximum allowed increase percentage for the recommended minimum group size, based on the existing minimum group size.
Use Case  
Values/Units Percentage of the existing minimum group size.
Dependencies  
Example  
## - Maximum Decrease
Description This setting defines the maximum allowed decrease percentage for the recommended minimum group size, based on the existing minimum group size.
Use Case  
Values/Units Percentage of the existing minimum group size.
Dependencies This setting cannot result in the group falling below the "Minimum Scale Group Size".
Example  
## Maximum Scale Group Size
Description This setting defines the maximum allowed number of EC2 instances to recommend for an ASG.
Use Case  
Values/Units An integer indicating the maximum group size.
Dependencies  
Example  
## - Headroom for Maximum Group Size
Description This setting defines the maximum group size increase percentage to recommend, on top of the existing maximum group size.
Use Case  
Values/Units Percentage of the "Maximum Scale Group Size" value.
Dependencies  
Example  
## - Maximum Increase
Description This setting defines the maximum allowed increase percentage for the recommended maximum group size, based on the existing maximum group size.
Use Case  
Values/Units Percentage of the existing maximum group size.
Dependencies  
Example  
## - Maximum Decrease
Description This setting defines the maximum allowed decrease percentage for the recommended maximum group size, based on the existing maximum group size.
Use Case  
Values/Units Percentage of the existing maximum group size.
Dependencies  
Example  
## Sizing Strategy
Description This setting defines the strategy to use when sizing EC2 or VM instances, within an ASG or VM Scale Set.
For "CPU and Memory" and "CPU only" options, the settings, listed below, indicated as "Custom..." are pre-defined in the extension. You can select "Custom" and then use the settings below to change the predefined values.
Use Case  
Values/Units Select "CPU and Memory", "CPU Only", or "Custom".
Dependencies If "CPU and Memory" or "CPU Only" is selected, the corresponding limits are based on the "Sizing Limit" setting below and the corresponding scoring strategy is based on the "Sizing Strategy" below.
If "Custom" is selected, then the advanced settings below, are used.
Example  
## - Custom Sizing Input Workloads You can overide some of these policy settings. Overrides are independent of the "Sizing Strategy" pick list; i.e. regardless of the value selected here the override, if provided will be applied when the analyses are run. See [Scale Group Overrides](../Densify_Com/Overriding_Cloud_Recommendations.htm#Scale_Group_Overrides) for details of options and for configuring overrides.
Description This setting defines the workload types to use for determining the optimal instance type for an ASG or VM Scale Set. This setting is used when "Custom" is selected for "Sizing Strategy".
Use Case  
Values/Units A colon-separated list of workload types.
Dependencies  
Example CPU\_CINT2006Rate\_Utilization;Raw\_Mem\_Utilization
## - Custom Sizing Output Workloads
Description This setting defines the workload types to use for displaying recommendation workload charts. This setting is used when "Custom" is selected for "Sizing Strategy".
Use Case  
Values/Units A colon-separated list of workload types.
Dependencies  
Example CPU\_Util\_Predicted\_Day; Mem\_Utilization\_As\_Pct
## - Custom Sizing Backfill Workloads \[CRB-53279] When analyzing the list of ASG candidates, Kubex uses the representative memory workload: ASG parent (ECS/EKS), busiest child, ASG, and workload days policy setting. If the ASG does not have memory data , the simulation does the following: * The ASG sizing extension uses memory constraints when running the simulation; * Fall back memory that is defined through this policy setting is used; * The simulation accounts for memory even if memory usage data is not available for ASG
Description This setting defines the workload types to use for backfilling resource utilization metrics. This setting is used when "Custom" is selected for "Sizing Strategy".
Use Case  
Values/Units A colon-separated list of workload types.
Dependencies The specified list of workloads must be the same as those defined for the -Custom Sizing Input Workloads.
Example Use the Transform\_CloudECS tag if available. It is published for ASGs that are children of ECS, otherwise use Mem\_Cloud
Mem Utilization As Pct, !Mem Cloud
## - Custom Sizing Backfill Parent Workloads
Description This setting defines the workload types to use for backfilling parent resource utilization metrics. This setting is used when "Custom" is selected for "Sizing Strategy".
Use Case  
Values/Units A colon-separated list of workload types.
Dependencies The specifed list of workloads must be the same as those defined for the -Custom Sizing Input Workloads.
Example Mem Utilization As Pct, !Transform\_CloudECS
## - Custom Sizing Scoring
Description This setting defines the scoring strategy to use when sizing EC2 or VM instances, within an ASG or VM Scale Set. This setting is used when "Custom" is selected for "Sizing Strategy".
Use Case  
Values/Units A colon-separated list of scoring strategies.
Dependencies  
Example  
## - Custom Sizing Limits
Description This setting defines the scoring limit to use when sizing EC2 or VM instances, within an ASG or VM Scale Set. This setting is used when 'Custom' is selected for 'Sizing Strategy'.
Use Case  
Values/Units A colon-separated list of scoring limits.
Dependencies  
Example 70;70
## Sizing Limit - Scoring Strategy
Description This setting defines the scoring strategy to use when sizing EC2 or VMs instances, within an ASG or VM Scale Set.
Use Case  
Values/Units Select one of the following scoring strategies:
  • Average
  • Peak
  • Sustained
See Scoring Strategy Setting on page 1 for details.
Dependencies This setting is used when either "CPU and Memory" or "CPU Only" is selected for "Sizing Strategy"
Example Sustained
## CPU Sizing Limit
Description This setting defines the CPU utilization limit to use when sizing EC2 or VM instances, within an ASG or VM Scale Set.
Use Case  
Values/Units A numeric value representing a percentage from 0 to 100.
Dependencies This setting is used when either "CPU and Memory" or "CPU Only" is selected for "Sizing Strategy".
Example  
## Memory Sizing Limit
Description This setting defines the memory utilization limit to use when sizing EC2 or VM instances, within an ASG or VM Scale Set.
Use Case  
Values/Units A numeric value representing a percentage from 0 to 100.
Dependencies This setting is used when "CPU and Memory" is selected for "Sizing Strategy".
Example  
## Worker Size Limit - Cores
Description This setting allows you to limit the increase of vCPU cores for the current size relative to recommended sizing, for ASG worker nodes. The range is typically 25 to 100%, with the higher percentages allowing a smaller increase. i.e. if set to 25%, the recommended worker node can have 4 times more vCPUs than the current worker node. A value of 100% allows no increase in vCPU allocation.

This setting is mapped as the low limit to Total CPUs and controls the maximum size of the recommended Instance type.
Use Case For ASG recommendations it is important to be able to control the maximum instance size in order to avoid using bigger (and more expensive) instances, even though it may appear to be optimal, large instances may introduce issues, if the ASG needs to scale up, rapidly.
Values/Units Size limit as a percentage. i.e. 25%
Dependencies  
Example  
## Worker Size Limit - Memory
Description This setting allows you to limit the increase the memory for the current size relative to recommended sizing, for ASG worker nodes. The range is typically 25 to 100%, with higher percentages allowing a smaller increase. i.e. if set to 25%, the recommended worker node can have 4 times more memory than the current worker node. A value of 100% allows no increase in memory allocation.

This setting is mapped as the low limit to Total Memory and controls the maximum size of the recommended Instance type.
Use Case For ASG recommendations it is important to be able to control the maximum instance size in order to avoid using bigger (and more expensive) instances, even though it may appear to be optimal, large instances may introduce issues, if the ASG needs to scale up, rapidly.
Values/Units Size limit as a percentage. i.e. 25%
Dependencies  
Example