New to KubeDB? Please start here.

QdrantAutoscaler

What is QdrantAutoscaler

QdrantAutoscaler is a Kubernetes Custom Resource Definitions (CRD). It provides a declarative configuration for automatic scaling of Qdrant compute resources (CPU, memory) and storage in a Kubernetes native way.

QdrantAutoscaler CRD Specifications

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

Sample QdrantAutoscaler for compute autoscaling:

apiVersion: autoscaling.kubedb.com/v1alpha1
kind: QdrantAutoscaler
metadata:
  name: qdrant-as-compute
  namespace: demo
spec:
  databaseRef:
    name: qdrant-sample
  opsRequestOptions:
    timeout: 3m
    apply: IfReady
  compute:
    node:
      trigger: "On"
      podLifeTimeThreshold: 10m
      resourceDiffPercentage: 20
      minAllowed:
        cpu: 400m
        memory: 400Mi
      maxAllowed:
        cpu: 1
        memory: 2Gi
      controlledResources: ["cpu", "memory"]
      containerControlledValues: "RequestsAndLimits"

Sample QdrantAutoscaler for storage autoscaling:

apiVersion: autoscaling.kubedb.com/v1alpha1
kind: QdrantAutoscaler
metadata:
  name: qdrant-as-storage
  namespace: demo
spec:
  databaseRef:
    name: qdrant-sample
  storage:
    node:
      trigger: "On"
      usageThreshold: 20
      scalingThreshold: 20
      expansionMode: "Online"

QdrantAutoscaler Spec

A QdrantAutoscaler object has the following fields in the spec section:

spec.databaseRef

spec.databaseRef is a required field that points to the Qdrant object for which autoscaling will be performed. It contains:

• spec.databaseRef.name — the name of the target Qdrant database (required).

spec.compute

spec.compute specifies the compute (CPU and memory) autoscaling configuration. It contains:

• spec.compute.node — the per-node compute autoscaling configuration:
  • trigger — enables (On) or disables (Off) compute autoscaling.
  • podLifeTimeThreshold — the minimum age of a pod before VPA can recommend resource updates.
  • resourceDiffPercentage — the minimum percentage difference required before applying a recommendation.
  • minAllowed — the minimum allowed CPU and memory resources.
  • maxAllowed — the maximum allowed CPU and memory resources.
  • controlledResources — the list of resources to be controlled (e.g., ["cpu", "memory"]).
  • containerControlledValues — specifies whether to control RequestsAndLimits or RequestsOnly.
  • inMemoryStorage — configuration for in-memory storage autoscaling per node:
    • scalingFactorPercentage — the scaling factor percentage for in-memory storage.
    • usageThresholdPercentage — the usage threshold percentage that triggers scaling.
• spec.compute.nodeTopology — specifies per-node topology for compute autoscaling:
  • name — the name of the topology entry.
  • scaleDownDiffPercentage — the scale-down difference percentage for this topology.
  • scaleUpDiffPercentage — the scale-up difference percentage for this topology.

spec.storage

spec.storage specifies the storage autoscaling configuration. It contains a node sub-section with the following fields:

• spec.storage.node.trigger — enables (On) or disables (Off) storage autoscaling.
• spec.storage.node.usageThreshold — the storage usage threshold (percentage) that triggers autoscaling.
• spec.storage.node.scalingThreshold — the percentage by which storage will be scaled when triggered.
• spec.storage.node.expansionMode — the volume expansion mode (Online or Offline).
• spec.storage.node.upperBound — the upper bound for storage size.
• spec.storage.node.scalingRules — a list of scaling rule objects, each containing:
  • appliesUpto — the upper limit for which this rule applies.
  • threshold — the threshold for this scaling rule.

spec.opsRequestOptions

spec.opsRequestOptions specifies the options for the QdrantOpsRequest created by the autoscaler. It contains:

• spec.opsRequestOptions.timeout — the timeout for the generated ops request.
• spec.opsRequestOptions.apply — when to apply the ops request. Can be Always or IfReady.
• spec.opsRequestOptions.maxRetries — the maximum number of retries for the ops request.

Next Steps

• Read the Qdrant autoscaler overview.
• See the compute autoscaler guide and storage autoscaler guide.