Create a Recommendation From a Policy Template

Flexera uses policies and their incident data to create recommendations which populate Cloud Cost Optimization’s Total Potential Savings chart to provide the Cloud Cost Optimization users insights on potential savings across their cloud estate and link them to Billing Centers for those owners to act.

The table in the following Policy Info Block subsection provides specific details to enable the policies to feed into the Total Potential Savings charts and data tables. There are two main changes:

• The meta data to define the policy as a recommendation, and,
• The incident standard fields that are read to populate the recommendations table.

Policy Info Block

Add the following fields to the info block as show in the example below.

Field

Example

Format

Required

Description

recommendation_type

Usage Reduction

string

yes

This populates the type field on the recommendation. Must be one of the known values: Rate Reduction or Usage Reduction.

provider

AWS

string

yes

This populates the vendor field on the recommendation.

policy_set

Unused Volumes

string

yes

This populates the Set field on the recommendation. Used to category like recommendations across cloud vendors

Example

name "AWS Unused Volumes"

rs_pt_ver 20180301

type "policy"

…

info(

  version: "3.6",

  service: "EBS",

  recommendation_type: “Usage Reduction”,

  policy_set: "Unused Volumes"

  provider: "AWS",

)

…

Standard Incident Fields

The fields are defined in the Policy Template policy declaration.

Field

Example

Format

Required

Description

savings

23.456

number

yes

Estimated monthly savings.

accountID

(ID capitalized)

"902149042019"

string

yes

AWS Account Number, Azure Subscription ID, or Google Project ID. Use to assign the recommendation to the billing center.

accountName

"Team1 Staging"

string

no

The user-friendly name for the above accountID. Used to assign the recommendation to the billing center.

resourceGroup

"My Resource Group"

string

no

Azure Resource Group. Use to assign the recommendation to the billing enter.

tags

[

  "env=dev",

  "tag2=value2"

]

 

array of strings

 

(each entry must use the "=" delimiter between tag key and value)

yes

The tags (labels from Google) on the resource. Use to assign the recommendation to the billing center.

resourceID

"db-3ZP2QSD2IL"

string

yes

The unique ID of the cloud vendor resource.

resourceType

"gp2"

string

no

May be a volume type, instance type, and so on.  

region

"us-east-1"

string

no

The cloud provider geo where the resource is provisioned.

service

"EC2"

string

no

The actual field populated on the incident, independent of a info:service object on the policy template.

term

"1 year"

string

no

Term of discount commitment.

platform

"Windows"

"MySQL"

string

no

Operating system, database engine.

paymentOption

"All Upfront"

string

no

Purchasing option of a discount commitment.

scope

"Shared"

string

no

Scope of a discount commitment.

licenseModel

"BYOL"

"License Included"

string

no

License model of a reservation or a resource.

deploymentOption

"Multi-AZ"

string

no

Deployment option of an RDS instance or RI.

averageUtilization

86

number

no

Predicted utilization for a recommended reservation.

lookbackPeriod

"30 days"

string

no

Time range analyzed the recommendation is based on.

resourceName

"testVM"

string

no

Resource name.

newResourceType

"r6.xlarge"

string

no

Recommended resource type if rightsizing is recommended.

size

4

number

no

Size of a resource. For example, volume size.

state

"unattached"

string

no

Resource state. For example, in-use for volume.

threshold

70

number

no

Tested threshold to produce recommendation.

thresholdType

"avg"

"p95"

string

no

Utilization metric percentile tested against threshold.

cpuAverage

27

number

no

CPU utilization average value.

cpuMaximum

89

number

no

CPU utilization maximum value.

cpuP95

56

number

no

CPU utilization 95th percentile value.

memAverage

25

number

no

Memory utilization average value.

memMaximum

80

number

no

Memory utilization maximum value.

memP95

77

number

no

Memory utilization 95th percentile value.

iopsAverage

100000

number

no

Input/Output operations per second average value.

recommendationDetails

“Terminate EC2 instance <instance ID> in AWS account <account name> (<account number>)”

string

no

The user-friendly description of the recommended action.

Example

policy "policy_unattached_volumes_list" do

…

    export "unused_volumes" do

      resource_level true

      field "accountID" do

        label "Account ID"

      end

      field "accountName" do

        label "Account Name"

      end

      field "region" do

        label "Region"

      end

      field "resourceID" do

        label "Resource ID"

        path "volumeId"

      end

      field "resourceType" do

        label "Resource Type"

      end

      field "size" do

        label "Size"

      end

      field "tags" do

        label "Tags"

      end

      field "savings" do

        label "Estimated Monthly Savings"

      end

      field "savingsCurrency" do

        label "Savings Currency"

      end

      field "id" do

        label "Id"

        path "volumeId"

      end

      field "service" do

        label "Service"

      end

      field "newResourceType" do

        label "Recommended Resource Type"

      end

      field "platform" do

        label "Platform"

      end

      field "cpuP95" do

        label "CPU Utilization (p95)"

        path "cpu_p95"

    end

  end

end

See the full public version of AWS Unused Volumes in our GitHub repository.