Azure Application Insights

Azure Application Insights ^{Click here for latest}

Scale applications based on Azure Application Insights metrics.

Availability: v2.6+ Maintainer: Microsoft

Warning

You are currently viewing v"2.11" of the documentation and it is not the latest. For the most recent documentation, kindly click here.

Trigger Specification

This specification describes the azure-app-insights trigger that scales based on an Azure Application Insights metric.

triggers:
- type: azure-app-insights
  metadata:
    metricAggregationTimespan: "0:1"
    metricAggregationType: avg
    metricFilter: cloud/roleName eq 'role_name'
    metricId: "customMetrics/example-metric"
    targetValue: "1.5"
    activationTargetValue: "5.5"
    activeDirectoryClientIdFromEnv: CLIENT_ID_ENV_NAME # Optional, can use TriggerAuthentication as well
    activeDirectoryClientPasswordFromEnv: CLIENT_PASSWORD_ENV_NAME # Optional, can use TriggerAuthentication as well
    applicationInsightsIdFromEnv: APP_ID # Optional, can use TriggerAuthentication as well
    tenantIdFromEnv: TENANT_ID` # Optional, can use TriggerAuthentication as well
    # Optional (Default: AzurePublicCloud)
    cloud: Private
    # Required when cloud = Private
    appInsightsResourceURL: https://api.applicationinsights.airgap.io/
    # Required when cloud = Private.
    activeDirectoryEndpoint: https://login.airgap.example/
    ignoreNullValues: true # Default is `false`. Set to `true` to ignore any errors with data extraction from a successful query. Set to `false` the scaler will return error when null values are discovered

This scaler is backed by the Azure Application Insights REST API. Please see this page for further details.

Parameter list:

tenantId - Id of the tenant that contains the Azure resource. This is used for authentication.
metricId - The name of the Application Insights metric to query. Use the Azure Command Line Interface to run az monitor app-insights metrics get-metadata to see a list of available metrics.
targetValue - Target value to trigger scaling actions. (This value can be a float)
activationTargetValue - Target value for activating the scaler. Learn more about activation here.(Default: 0, Optional, This value can be a float)
metricAggregationType - Aggregation method of the Azure Application Insights metric. The aggregation methods vary from metric to metric. The az monitor app-insights metrics get-metadata command can be used to determine which methods apply to a given metric. (Some common aggregation methods are avg, count, sum, min, and max)
metricAggregationInterval - Collection time of the metric in format "hh:mm".
applicationInsightsId - Id of the Application Insights instance to query. This is a GUID that can be retrieved from the Application Insight’s API Access blade in the Azure Portal.
activeDirectoryClientId - Id of the Active Directory client. The client must have Monitoring Reader permissions for the Application Insights instance.
activeDirectoryClientPassword - Password of the Active Directory client password.
metricFilter - Further specify the metrics query using a filter. For example cloud/roleName eq 'example. (Optional)
cloud - Name of the cloud environment that the Azure Application Insights instance belongs to. (Values: AzurePublicCloud, AzureUSGovernmentCloud, AzureChinaCloud, Private, Default: AzurePublicCloud, Optional)
appInsightsResourceURL - Application Insights REST API URL of the cloud environment. (Required when cloud is set to Private, e.g. https://api.applicationinsights.azure.cn/ for AzureChinaCloud).
activeDirectoryEndpoint - Active Directory endpoint of the cloud environment. (Required when cloud is set to Private, e.g. https://login.chinacloudapi.cn/ for AzureChinaCloud).
ignoreNullValues - Ignore any errors with data extraction from a successful query (Values: true,false, Default: false, Optional).

Some parameters can be provided using environment variables, instead of setting them directly in metadata. Here is a list of parameters you can use to retrieve values from environment variables:

activeDirectoryClientIdFromEnv - Name of the environment variable that contains the Id of the Active Directory application. (Optional)
activeDirectoryClientPasswordFromEnv - Name of the environment variable that contains the Active Directory client password. (Optional)
applicationInsightsIdFromEnv - Name of the environment variable that contains the Application Insights Id. (Optional)
tenantIdFromEnv - Name of the environment variable that contains the Id of the tenant that contains the Application Insights instance. (Optional)

Authentication Parameters

You can use the TriggerAuthentication CRD to configure the authentication by providing a set of Azure Active Directory credentials or by using pod identity.

Credential based authentication:

activeDirectoryClientId - Id of the Active Directory application which requires at least Monitoring Reader permissions.
activeDirectoryClientPassword - Password of the Active Directory application.
applicationInsightsId - Id of the Application Insights instance to query.
tenantId - Id of the tenant that contains the Azure resource.

The principal will need Monitoring Reader access to query metrics from the Application Insights instance.

Pod identity based authentication:

Azure AD Pod Identity or Azure AD Workload Identity providers can be used in place of credential based authentication. The following section contains an example of a TriggerAuthentication using pod identity.

Example

The following example illustrates the use of a TriggerAuthentication to connect to Application Insights.

apiVersion: v1
kind: Secret
metadata:
  name: azure-app-insights-secrets
data:
  activeDirectoryClientId: <clientId>
  activeDirectoryClientPassword: <clientPassword>
  applicationInsightsId: <appInsightsAppId>
  tenantId: <tenantId>
---
apiVersion: keda.sh/v1alpha1
kind: TriggerAuthentication
metadata:
  name: azure-app-insights-trigger-auth
spec:
  secretTargetRef:
    - parameter: activeDirectoryClientId
      name: azure-app-insights-secrets
      key: activeDirectoryClientId
    - parameter: activeDirectoryClientPassword
      name: azure-app-insights-secrets
      key: activeDirectoryClientPassword
    - parameter: applicationInsightsId
      name: azure-app-insights-secrets
      key: applicationInsightsId
    - parameter: tenantId
      name: azure-app-insights-secrets
      key: tenantId
  # or Pod Identity, kind: Secret is not required in case of pod Identity
  podIdentity:
      provider: azure | azure-workload
---
apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: azure-app-insights-scaler
spec:
  scaleTargetRef:
    name: azure-app-insights-example
  minReplicaCount: 1
  maxReplicaCount: 10
  triggers:
  - type: azure-app-insights
    metadata:
      metricId: "customMetrics/example-metric"
      metricAggregationTimespan: "0:5"
      metricAggregationType: avg
      metricFilter: cloud/roleName eq 'example'
      targetValue: "1"
    authenticationRef:
      name: azure-app-insights-trigger-auth

The following example illustrates the use of environment variables to connect to Application Insights.

apiVersion: v1
kind: Secret
metadata:
  name: azure-app-insights-secrets
type: Opaque
data:
  activeDirectoryClientId: <clientId>
  activeDirectoryClientPassword: <clientPassword>
  applicationInsightsId: <appInsightsAppId>
  tenantId: <tenantId>
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: azure-app-insights-example
spec:
  replicas: 0
  selector:
    matchLabels:
      app: azure-app-insights-example
  template:
    metadata:
      labels:
        app: azure-app-insights-example
    spec:
      containers:
      - name: example
        image: nginx:1.16.1
        env:
        - name: ACTIVE_DIRECTORY_ID
          valueFrom:
            secretKeyRef:
              name: azure-app-insights-secrets
              key: activeDirectoryClientId
        - name: ACTIVE_DIRECTORY_PASSWORD
          valueFrom:
            secretKeyRef:
              name: azure-app-insights-secrets
              key: activeDirectoryClientPassword
        - name: APP_INSIGHTS_APP_ID
          valueFrom:
            secretKeyRef:
              name: azure-app-insights-secrets
              key: applicationInsightsId
        - name: TENANT_ID
          valueFrom:
            secretKeyRef:
              name: azure-app-insights-secrets
              key: tenantId
---
apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: azure-app-insights-scaler
spec:
  scaleTargetRef:
    name: azure-app-insights-example
  pollingInterval: 5
  cooldownPeriod: 5
  minReplicaCount: 0
  maxReplicaCount: 2
  triggers:
    - type: azure-app-insights
      metadata:
        metricId: "customMetrics/example-metric"
        metricAggregationTimespan: "0:5"
        metricAggregationType: avg
        metricFilter: cloud/roleName eq 'example'
        targetValue: "1"
        activeDirectoryClientIdFromEnv: ACTIVE_DIRECTORY_ID
        activeDirectoryClientPasswordFromEnv: ACTIVE_DIRECTORY_PASSWORD
        applicationInsightsIdFromEnv: APP_INSIGHTS_APP_ID
        tenantIdFromEnv: TENANT_ID