Kubernetes使用Keda进行弹性伸缩，更合理利用资源 java

我最新最全的文章都在南瓜慢说 www.pkslow.com ，欢迎大家来喝茶！

1 简介 Kubernetes自带的HPA是只支持CPU/MEM的，很多时候我们并不根据这两项指标来进行伸缩资源。比如消费者不断处理MQ的消息，我们希望MQ如果堆积过多，就启动更多的消费者来处理任务。而Keda给了我们很多选择。

文章图片

KEDA 是 Kubernetes 基于事件驱动的自动伸缩工具，通过 KEDA 我们可以根据需要处理的事件数量来驱动 Kubernetes 中任何容器的扩展。KEDA 可以直接部署到任何 Kubernetes 集群中和标准的组件一起工作。
Keda所支持的事件源非常丰富，本文我们以RabbitMQ为例进行演示。

文章图片

2 安装Keda 安装的方法很多，我们直接通过yaml文件来安装，这样还可以修改镜像地址等。先从（ https://github.com/kedacore/k... ）下载yaml文件，然后执行：

$ kubectl apply -f ~/Downloads/keda-2.2.0.yaml namespace/keda created customresourcedefinition.apiextensions.k8s.io/clustertriggerauthentications.keda.sh created customresourcedefinition.apiextensions.k8s.io/scaledjobs.keda.sh created customresourcedefinition.apiextensions.k8s.io/scaledobjects.keda.sh created customresourcedefinition.apiextensions.k8s.io/triggerauthentications.keda.sh created serviceaccount/keda-operator created clusterrole.rbac.authorization.k8s.io/keda-external-metrics-reader created clusterrole.rbac.authorization.k8s.io/keda-operator created rolebinding.rbac.authorization.k8s.io/keda-auth-reader created clusterrolebinding.rbac.authorization.k8s.io/keda-hpa-controller-external-metrics created clusterrolebinding.rbac.authorization.k8s.io/keda-operator created clusterrolebinding.rbac.authorization.k8s.io/keda:system:auth-delegator created service/keda-metrics-apiserver created deployment.apps/keda-metrics-apiserver created deployment.apps/keda-operator created apiservice.apiregistration.k8s.io/v1beta1.external.metrics.k8s.io created

检查一下是否都已经启动完成：

$ kubectl get all -n keda NAMEREADYSTATUSRESTARTSAGE pod/keda-metrics-apiserver-55dc9f9498-smc2d1/1Running02m41s pod/keda-operator-59dcf989d6-pxcbb1/1Running02m41sNAMETYPECLUSTER-IPEXTERNAL-IPPORT(S)AGE service/keda-metrics-apiserverClusterIP10.104.255.44443/TCP,80/TCP2m41sNAMEREADYUP-TO-DATEAVAILABLEAGE deployment.apps/keda-metrics-apiserver1/1112m42s deployment.apps/keda-operator1/1112m42sNAMEDESIREDCURRENTREADYAGE replicaset.apps/keda-metrics-apiserver-55dc9f94981112m42s replicaset.apps/keda-operator-59dcf989d61112m42s

也可以看到镜像多了：

$ docker images | grep keda ghcr.io/kedacore/keda-metrics-apiserver2.2.0a43d404533686 weeks ago95.3MB ghcr.io/kedacore/keda2.2.042b88f0429146 weeks ago83MB

如果要卸载请执行：

$ kubectl delete -f ~/Downloads/keda-2.2.0.yaml

3 安装RabbitMQ 为了快速安装，也方便日后删除，我们通过Helm来安装RabbitMQ。
查看可用的chart：

$ helm search repo rabbit

执行安装：

$ helm install azure-rabbitmq azure/rabbitmq

检查一下：

$ helm list NAMENAMESPACEREVISIONUPDATEDSTATUSCHARTAPP VERSION azure-ingressdefault12021-02-14 01:21:07.212107 +0800 CSTdeployednginx-ingress-1.41.3v0.34.1 azure-rabbitmqdefault12021-05-05 11:29:06.979437 +0800 CSTdeployedrabbitmq-6.18.23.8.2

用户名为user，密码获取如下：

$ echo "Password: $(kubectl get secret --namespace default azure-rabbitmq -o jsonpath="{.data.rabbitmq-password}" | base64 --decode)" Password: YNsEayx8w2

4 测试部署消费者，注意这里有个MQ连接信息和加密，要根据自己情况修改。

$ kubectl apply -f src/main/kubernetes/deploy-consumer.yaml secret/rabbitmq-consumer-secret created deployment.apps/rabbitmq-consumer created scaledobject.keda.sh/rabbitmq-consumer created triggerauthentication.keda.sh/rabbitmq-consumer-trigger created

查看deployment，发现是没有Pod创建，因为还不需要处理，MQ现在的队列为0。

$ kubectl get deployments NAMEREADYUP-TO-DATEAVAILABLEAGE azure-ingress-nginx-ingress-controller1/11180d azure-ingress-nginx-ingress-default-backend1/11180d rabbitmq-consumer0/000131m

部署生产者，往MQ发送消息：

$ kubectl apply -f src/main/kubernetes/deploy-publisher-job.yaml job.batch/rabbitmq-publish created

可以看到，慢慢消费者就起来了，并且创建了越来越多的Pod来处理MQ：

$ kubectl get deployments rabbitmq-consumer NAMEREADYUP-TO-DATEAVAILABLEAGE rabbitmq-consumer1/111167m$ kubectl get deployments rabbitmq-consumer NAMEREADYUP-TO-DATEAVAILABLEAGE rabbitmq-consumer3/443168m$ kubectl get deployments rabbitmq-consumer NAMEREADYUP-TO-DATEAVAILABLEAGE rabbitmq-consumer4/884168m$ kubectl get deployments rabbitmq-consumer NAMEREADYUP-TO-DATEAVAILABLEAGE rabbitmq-consumer6/886169m $ kubectl get deployments rabbitmq-consumer NAMEREADYUP-TO-DATEAVAILABLEAGE rabbitmq-consumer0/000171m

查看Deployment的Event也可以看到结果：

Events: TypeReasonAgeFromMessage ------------------------- NormalScalingReplicaSet5m55s (x2 over 172m)deployment-controllerScaled up replica set rabbitmq-consumer-7b477f78b4 to 1 NormalScalingReplicaSet5m6sdeployment-controllerScaled up replica set rabbitmq-consumer-7b477f78b4 to 4 NormalScalingReplicaSet4m6sdeployment-controllerScaled up replica set rabbitmq-consumer-7b477f78b4 to 8 NormalScalingReplicaSet3m5sdeployment-controllerScaled up replica set rabbitmq-consumer-7b477f78b4 to 16 NormalScalingReplicaSet3m3s (x2 over 172m)deployment-controllerScaled down replica set rabbitmq-consumer-7b477f78b4 to 0

处理完成后，又会回到0了。
总结代码请查看：https://github.com/LarryDpk/p...
欢迎关注微信公众号<南瓜慢说>，将持续为你更新...