Kedify | CouchDB

Use events from CouchDB to trigger autoscaling with Kedify and KEDA

CouchDB is a NoSQL database well-suited for web applications, using JSON for documents and featuring a robust, user-friendly HTTP API.

Integrated with KEDA, CouchDB enables precise autoscaling of Kubernetes applications based on real-time data changes, allowing for highly responsive and efficient scaling strategies tailored to current workload demands.

Featured Use Cases

Scenario:

A content management system needs to scale its caching layer dynamically as the number of new comments or page views spikes during high traffic events.

CouchDB Scaler Usage:

Scaling is triggered based on the count of documents tagged with "new: true" in the interactions database.

KEDA Usage:

The CMS uses KEDA to query the count of new interactions, automatically adjusting the number of cache instances to handle increased load, ensuring optimal performance and user experience.

Get Started

                    apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: cms-cache-scaler
  namespace: default
spec:
  scaleTargetRef:
    name: cache-service
  minReplicaCount: 2
  maxReplicaCount: 15
  triggers:
  - type: couchdb
    metadata:
      host: 'couchdb-service.default.svc.cluster.local'
      port: '5984'
      dbName: 'cms_interactions'
      query: '{ "selector": { "isNew": true }, "fields": ["_id"] }'
      queryValue: '100'
      activationQueryValue: '50'

Scenario:

Financial institutions trigger batch jobs to perform fraud analysis when the number of transactions flagged for review exceeds a certain threshold.

CouchDB Scaler Usage:

Scaling occurs based on the count of transactions marked as "reviewNeeded: true" .

KEDA Usage:

The financial service uses KEDA to scale out analysis jobs, allowing for rapid response to potential fraud, which helps in reducing financial losses and maintaining customer trust.

Get Started

                    apiVersion: keda.sh/v1alpha1
kind: ScaledJob
metadata:
  name: fraud-detection-job
  namespace: default
spec:
  jobTargetRef:
    template:
      spec:
        containers:
        - name: fraud-analysis-worker
          image: fraud-detection-algorithm
          env:
          - name: DB_HOST
            value: 'couchdb.default.svc.cluster.local'
          - name: DB_PORT
            value: '5984'
          - name: DB_NAME
            value: 'transactions_db'
          command:
          - 'python'
          - '-u'
          - 'analyze.py'
  triggers:
  - type: couchdb
    metadata:
      host: 'couchdb.default.svc.cluster.local'
      port: '5984'
      dbName: 'transactions_db'
      query: '{ "selector": { "reviewNeeded": true }, "fields": ["_id", "amount"] }'
      queryValue: '30'
      activationQueryValue: '5'

Scenario:

A social media platform scales its data processing service to analyze trends based on the volume of new posts and comments on trending topics.

CouchDB Scaler Usage:

Scaling is based on the number of posts and comments containing specific trending hashtags.

KEDA Usage:

KEDA enables the platform to automatically adjust the scale of its analytics services as social interactions peak, ensuring timely processing of trend data for user engagement.

Get Started

                    apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: social-media-trend-scaler
  namespace: default
spec:
  scaleTargetRef:
    name: trend-analysis-service
  minReplicaCount: 1
  maxReplicaCount: 20
  triggers:
  - type: couchdb
    metadata:
      host: 'couchdb.default.svc.cluster.local'
      port: '5984'
      dbName: 'social_media'
      query: '{ "selector": { "hashtag": "#trending" }, "fields": ["_id"] }'
      queryValue: '200'
      activationQueryValue: '100'