Home >Backend Development >Golang >Analyzing Kubernetes gRPC load balancing (L4 vs L7)
This article is introduced by the go language tutorial column to introduce gRPC load balancing in Kubernetes. I hope it will be helpful to friends in need!
Download and install Docker Desktop and start the built-in Kubernetes cluster.
# 安装 Gobrew install go# 安装 Protobufbrew install protobuf# 安装 Istiobrew install istioctl kubectl config use-context docker-desktop istioctl install -y
github.com/jxlwqq/grpc-lb
Pull code:
git clone git@github.com:jxlwqq/grpc-lb.gitcd grpc-lb
Command | Description |
---|---|
##make init
| Install protoc-gen-go and protoc-gen-grpc|
make protoc
| Based on proto File, generate *_pb.go and *_grpc.pb.go|
make docker-build
| Build docker image|
make kube-deploy
| Deploy services in the cluster|
Delete service |
|
Inject Istio sidecar |
Envoy (Istio) is more versatile and supports all HTTP/2 functions requested and responded by gRPC as the underlying routing and load balancing.
make docker-build # 构建镜像(构建好的镜像,不 push 到远程仓库中)
View the image:
docker images ls
Return:
REPOSITORY TAG IMAGE ID CREATED SIZE grpc-lb/client-grpc latest 95d32ead8d9b 12 seconds ago 16.6MB grpc-lb/client-http latest dbf0341206f6 22 seconds ago 11.5MB grpc-lb/server latest 1ef346785b2a 29 seconds ago 18.2MB
Deployment Go to the cluster:
make kube-deploy # 在集群中部署服务
View Pod:
kubectl get pods
Return:
NAME READY STATUS RESTARTS AGE client-grpc-6c565594f4-tdf75 1/1 Running 0 2m48s client-http-55d95c744d-f7nx4 1/1 Running 0 2m49s server-7c4bfd74d-29c69 1/1 Running 0 2m51s server-7c4bfd74d-4btvw 1/1 Running 0 2m51s server-7c4bfd74d-fk8zf 1/1 Running 0 2m51s
View the log of client-http Pod:
export CLIENT_HTTP_POD=$(kubectl get pod -l app=client-http -o jsonpath={.items..metadata.name})kubectl logs "${CLIENT_HTTP_POD}"
Return:
#1: server-7c4bfd74d-4btvw#2: server-7c4bfd74d-4btvw#3: server-7c4bfd74d-29c69#4: server-7c4bfd74d-fk8zf#5: server-7c4bfd74d-fk8zf#6: server-7c4bfd74d-29c69#7: server-7c4bfd74d-fk8zf#8: server-7c4bfd74d-4btvw#9: server-7c4bfd74d-fk8zf
View the log of client-grpc Pod:
export CLIENT_GRPC_POD=$(kubectl get pod -l app=client-grpc -o jsonpath={.items..metadata.name})kubectl logs "${CLIENT_GRPC_POD}"
Return:
#1: server-7c4bfd74d-fk8zf#2: server-7c4bfd74d-fk8zf#3: server-7c4bfd74d-fk8zf#4: server-7c4bfd74d-fk8zf#5: server-7c4bfd74d-fk8zf#6: server-7c4bfd74d-fk8zf#7: server-7c4bfd74d-fk8zf#8: server-7c4bfd74d-fk8zf#9: server-7c4bfd74d-fk8zf
It can be seen that the HTTP request is carrying a payload, while the RPC request is carrying an invalid load.
Manual injection:
make istio-inject # 注入 Istio 边车
View Pod:
kubectl get pods
Return:
NAME READY STATUS RESTARTS AGE client-grpc-7864f57779-f6blx 2/2 Running 0 17s client-http-f8964854c-jclkd 2/2 Running 0 21s server-7846bd6bb4-bcfws 2/2 Running 0 27s server-7846bd6bb4-fv29s 2/2 Running 0 40s server-7846bd6bb4-hzqj6 2/2 Running 0 34s
View the log of client-http Pod:
export CLIENT_HTTP_POD=$(kubectl get pod -l app=client-http -o jsonpath={.items..metadata.name})kubectl logs "${CLIENT_HTTP_POD}"
Return:
#1: server-7846bd6bb4-hzqj6#2: server-7846bd6bb4-fv29s#3: server-7846bd6bb4-hzqj6#4: server-7846bd6bb4-hzqj6#5: server-7846bd6bb4-hzqj6#6: server-7846bd6bb4-hzqj6#7: server-7846bd6bb4-hzqj6#8: server-7846bd6bb4-bcfws#9: server-7846bd6bb4-fv29s
View the log of client-grpc Pod:
export CLIENT_GRPC_POD=$(kubectl get pod -l app=client-grpc -o jsonpath={.items..metadata.name})kubectl logs "${CLIENT_GRPC_POD}"
Return:
#1: server-7846bd6bb4-fv29s#2: server-7846bd6bb4-hzqj6#3: server-7846bd6bb4-fv29s#4: server-7846bd6bb4-bcfws#5: server-7846bd6bb4-fv29s#6: server-7846bd6bb4-hzqj6#7: server-7846bd6bb4-fv29s#8: server-7846bd6bb4-bcfws#9: server-7846bd6bb4-fv29s
It can be seen that both HTTP requests and RPC requests are carrying payloads.
make kube-delete istioctl experimental uninstall --purge
The above is the detailed content of Analyzing Kubernetes gRPC load balancing (L4 vs L7). For more information, please follow other related articles on the PHP Chinese website!