A simple docker container that will receive messages from a RabbitMQ queue and scale via KEDA. The reciever will receive a single message at a time (per instance), and sleep for 1 second to simulate performing work. When adding a massive amount of queue messages, KEDA will drive the container to scale out according to the event source (RabbitMQ).
- Kubernetes cluster
- KEDA installed on the cluster
This setup will go through creating a RabbitMQ queue on the cluster and deploying this consumer with the ScaledObject
to scale via KEDA. If you already have RabbitMQ you can use your existing queues.
First you should clone the project:
git clone https://github.com/GokGokalp/sample-netcore-keda-rabbitmq
cd sample-netcore-keda-rabbitmq
helm install --name rabbitmq --set rabbitmq.username=user,rabbitmq.password=PASSWORD stable/rabbitmq
kubectl get po
NAME READY STATUS RESTARTS AGE
rabbitmq-0 1/1 Running 0 3m3s
kubectl apply -f deploy/deploy-consumer.yaml
kubectl get deploy
You should see rabbitmq-consumer
deployment with 0 pods as there currently aren't any queue messages. It is scale to zero.
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
rabbitmq-consumer 0 0 0 0 3s
This consumer is set to consume one message per instance, sleep for 1 second, and then acknowledge completion of the message. This is used to simulate work. The ScaledObject
included in the above deployment is set to scale to a minimum of 0 replicas on no events, and up to a maximum of 30 replicas on heavy events (optimizing for a queue length of 5 message per replica). After 30 seconds of no events the replicas will be scaled down (cooldown period). These settings can be changed on the ScaledObject
as needed.
The following job will publish 300 messages to the "hello" queue the deployment is listening to. As the queue builds up, KEDA will help the horizontal pod autoscaler add more and more pods until the queue is drained after about 2 minutes and up to 30 concurrent pods. You can modify the exact number of published messages in the deploy-publisher-job.yaml
file.
kubectl apply -f deploy/deploy-publisher-job.yaml
kubectl get deploy -w
You can watch the pods spin up and start to process queue messages. As the message length continues to increase, more pods will be pro-actively added.
You can see the number of messages vs the target per pod as well:
kubectl get hpa
After the queue is empty and the specified cooldown period (a property of the ScaledObject
, default of 300 seconds) the last replica will scale back down to zero.
kubectl delete job rabbitmq-publish
kubectl delete ScaledObject rabbitmq-consumer
kubectl delete deploy rabbitmq-consumer
helm delete rabbitmq