Auto-scaling containers with KEDA

Slide deck of the presentation done at Credit Agricole Corporate and Investment Bank demonstrating KEDA capabilities. The talk focused on different options for scaling in Kubernetes cluster. The demo covered the auto scaling options based on events using KEDA project.

Transcript

1. Auto-Scaling containers with Nilesh Gule @nileshgule | www.HandsOnArchitect.com
2. $whoami { “name” : “Nilesh Gule”, “website” : “https://www.HandsOnArchitect.com", “github” : “https://github.com/NileshGule" “twitter” : “@nileshgule”, “linkedin” : “https://www.linkedin.com/in/nileshgule”, “email” : “nileshgule@gmail.com", “likes” : “Technical Evangelism, Cricket” }
6. Pros Cons Scale k8s objects: Kubectl scale –replicas=3 processor-deployment.yml Kubectl scale –replicas=3 deployment/rabbitmq-consumer Scale AKS cluster: az aks scale --resource-group myResourceGroup \ --name myAKSCluster \ --node-count 5 \ --nodepool-name <your node pool name>
7. Pros Cons kubectl autoscale deployment php-apache \ --cpu-percent=50 \ --min=1 \ --max=10
9. Pros Cons
10. Pros Cons
11. Kubernetes-based Event Driven Autoscaling
14. Azure Event Hub Azure Service Bus Queues and Topics Azure Storage Queues Kafka Prometheus RabbitMQ Redis Lists Liiklus NATS AWS CloudWatch AWS Simple Queue Service GCP PubSub Integrates with Azure functions
15. Demo Autoscaling Storage Queue consumer with ACI
16. 5000 messages
17. apiVersion: keda.k8s.io/v1alpha1 kind: ScaledObject metadata: name: hello-keda namespace: default labels: deploymentName: hello-keda spec: scaleTargetRef: deploymentName: hello-keda pollingInterval: 5 # Optional. Default: 30 seconds maxReplicaCount: 15 # Optional. Default: 100 triggers: - type: azure-queue metadata: name: myQueueItem type: queueTrigger direction: in queueName: js-queue-items connection: AzureWebJobsStorage
18. Demo Autoscaling RabbitMQ
19. https://github.com/NileshGule/AKS-learning-series
20. 1000 messages Batch of 10 messages
21. apiVersion: keda.k8s.io/v1alpha1 kind: ScaledObject metadata: name: rabbitmq-consumer namespace: default labels: deploymentName: rabbitmq-consumer spec: scaleTargetRef: deploymentName: rabbitmq-consumer pollingInterval: 5 # Optional. Default: 30 seconds cooldownPeriod: 30 # Optional. Default: 300 seconds minReplicaCount: 1 # Optional. Default: 0 maxReplicaCount: 30 # Optional. Default: 100 triggers: - type: rabbitmq metadata: queueName: hello host: 'amqp://user:PASSWORD@rabbitmq.default.svc.cluster.local:5672’ queueLength : '5'
22. AKS cluster-autoscalar Virtual Node Horizontal pod autoscale virtual kubelet KEDA Azure Function with Queue
25. Thank you very much Code with Passion and Strive for Excellence https://github.com/NileshGule/AKS-learning-series https://github.com/NileshGule/pd-tech-fest-2019 https://github.com/NileshGule/sample-dotnet-core-rabbitmq-keda https://www.slideshare.net/nileshgule/presentations https://speakerdeck.com/nileshgule/
26. Q&A