New to KubeDB? Please start here.

DruidAutoscaler

What is DruidAutoscaler

DruidAutoscaler is a Kubernetes Custom Resource Definitions (CRD). It provides a declarative configuration for autoscaling Druid compute resources and storage of database components in a Kubernetes native way.

DruidAutoscaler CRD Specifications

Like any official Kubernetes resource, a DruidAutoscaler has TypeMeta, ObjectMeta, Spec and Status sections.

Here, some sample DruidAutoscaler CROs for autoscaling different components of database is given below:

Sample DruidAutoscaler for druid cluster:

apiVersion: autoscaling.kubedb.com/v1alpha1
kind: DruidAutoscaler
metadata:
  name: dr-autoscaler
  namespace: demo
spec:
  databaseRef:
    name: druid-prod
  opsRequestOptions:
    timeout: 3m
    apply: IfReady
  compute:
    coordinators:
      trigger: "On"
      podLifeTimeThreshold: 24h
      minAllowed:
        cpu: 200m
        memory: 300Mi
      maxAllowed:
        cpu: 1
        memory: 1Gi
      controlledResources: ["cpu", "memory"]
      containerControlledValues: "RequestsAndLimits"
      resourceDiffPercentage: 10
    brokers:
      trigger: "On"
      podLifeTimeThreshold: 24h
      minAllowed:
        cpu: 200m
        memory: 300Mi
      maxAllowed:
        cpu: 1
        memory: 1Gi
      controlledResources: ["cpu", "memory"]
      containerControlledValues: "RequestsAndLimits"
      resourceDiffPercentage: 10
  storage:
    historicals:
      expansionMode: "Online"
      trigger: "On"
      usageThreshold: 60
      scalingThreshold: 50
    middleManagers:
      expansionMode: "Online"
      trigger: "On"
      usageThreshold: 60
      scalingThreshold: 50

Here, we are going to describe the various sections of a DruidAutoscaler crd.

A DruidAutoscaler object has the following fields in the spec section.

spec.databaseRef

spec.databaseRef is a required field that point to the Druid object for which the autoscaling will be performed. This field consists of the following sub-field:

  • spec.databaseRef.name : specifies the name of the Druid object.

spec.opsRequestOptions

These are the options to pass in the internally created opsRequest CRO. opsRequestOptions has two fields.

spec.compute

spec.compute specifies the autoscaling configuration for the compute resources i.e. cpu and memory of the database components. This field consists of the following sub-field:

  • spec.compute.coordinators indicates the desired compute autoscaling configuration for coordinators of a topology Druid database.
  • spec.compute.overlords indicates the desired compute autoscaling configuration for overlords of a topology Druid database.
  • spec.compute.brokers indicates the desired compute autoscaling configuration for brokers of a topology Druid database.
  • spec.compute.routers indicates the desired compute autoscaling configuration for routers of a topology Druid database.
  • spec.compute.historicals indicates the desired compute autoscaling configuration for historicals of a topology Druid database.
  • spec.compute.middleManagers indicates the desired compute autoscaling configuration for middleManagers of a topology Druid database.

All of them has the following sub-fields:

  • trigger indicates if compute autoscaling is enabled for this component of the database. If “On” then compute autoscaling is enabled. If “Off” then compute autoscaling is disabled.
  • minAllowed specifies the minimal amount of resources that will be recommended, default is no minimum.
  • maxAllowed specifies the maximum amount of resources that will be recommended, default is no maximum.
  • controlledResources specifies which type of compute resources (cpu and memory) are allowed for autoscaling. Allowed values are “cpu” and “memory”.
  • containerControlledValues specifies which resource values should be controlled. Allowed values are “RequestsAndLimits” and “RequestsOnly”.
  • resourceDiffPercentage specifies the minimum resource difference between recommended value and the current value in percentage. If the difference percentage is greater than this value than autoscaling will be triggered.
  • podLifeTimeThreshold specifies the minimum pod lifetime of at least one of the pods before triggering autoscaling.

There are two more fields, those are only specifiable for the percona variant inMemory databases.

  • inMemoryStorage.UsageThresholdPercentage If db uses more than usageThresholdPercentage of the total memory, memoryStorage should be increased.
  • inMemoryStorage.ScalingFactorPercentage If db uses more than usageThresholdPercentage of the total memory, memoryStorage should be increased by this given scaling percentage.

spec.storage

spec.storage specifies the autoscaling configuration for the storage resources of the database components. This field consists of the following sub-field:

  • spec.storage.historicals indicates the desired storage autoscaling configuration for historicals of a topology Druid cluster.
  • spec.storage.middleManagers indicates the desired storage autoscaling configuration for middleManagers of a topology Druid cluster.

spec.storage is only supported for druid data nodes i.e. historicals and middleManagers as they are the only nodes containing volumes.

All of them has the following sub-fields:

  • trigger indicates if storage autoscaling is enabled for this component of the database. If “On” then storage autoscaling is enabled. If “Off” then storage autoscaling is disabled.
  • usageThreshold indicates usage percentage threshold, if the current storage usage exceeds then storage autoscaling will be triggered.
  • scalingThreshold indicates the percentage of the current storage that will be scaled.
  • expansionMode indicates the volume expansion mode.