Cypress自动化Github workflow

目录

Workflow

.github/workflows/build.yaml

jobs

build-deploy:

.github/wrokflows/execute-tests.yaml

输入参数

jobs

restart_failed_dp:

seed_data:

execute-tests:

Docker File

docker/Dockerfile.base

docker/service-dockerfile.base

Deploy

deploy/test-execute

---

Workflow

.github/workflows/build.yaml

创建docker image并推送到Amazon ECR

jobs

build-deploy:

runs-on: self-hostedsteps:- name: "Checkout source code"uses: actions/checkout@v2- name: "Prepare environment"id: preprun: |# Source our global environment variables & add them to GITHUB_ENV for subsequent stepssource .github/global.envcat .github/global.env >> ${GITHUB_ENV}# Configure our environment based on the branch, or if this is a PR buildif [[ '${{ github.ref }}' == refs/heads/${MAIN_BRANCH} ]]; thenecho "::set-output name=tag::${{ github.sha }}"source .github/auto.envcat .github/auto.env >> ${GITHUB_ENV}elseecho "This isn't either the main branch or a pull request. Exiting..."exit 1fiecho "::set-output name=kubernetes-namespace::${KUBERNETES_NAMESPACE}"- name: Build, tag, and push image to Amazon ECRenv:IMAGE_TAG: ${{ steps.prep.outputs.tag }}GITHUB_TOKEN: ${{ secrets.ORG_TOKEN }}run: |# This is needed because the Github Action environment variables have issues if the value contains spaces (e.g. 'EXTRA_DOCKER_ARGS')source .github/global.env# Build and push the image using the SHA tagdocker build -f ${{ env.DOCKERFILE }} -t ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME }}:${IMAGE_TAG} ${EXTRA_DOCKER_ARGS} .docker push ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME }}:${IMAGE_TAG}docker build -f ${{ env.SERVICE_DOCKERFILE }} -t ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME_SERVICE }}:latest .docker push ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME_SERVICE }}:latest# Push this image with the 'latest' tag if this is a "main branch" buildif [[ '${{ github.ref }}' == 'refs/heads/${{ env.MAIN_BRANCH }}' ]]; thendocker tag ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME }}:${IMAGE_TAG} ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME }}:latestdocker push ${{ env.ECR_REGISTRY }}/${{ env.APP_NAME }}:latestfi

备注：

以上脚本执行了下列命令

docker build -f docker/Dockerfile.base -t 590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation:latest ${EXTRA_DOCKER_ARGS} .
docker push 590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation:latest

docker build -f docker/service-dockerfile.base -t 590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation-service:latest .
docker push 590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation-service:latest

.github/wrokflows/execute-tests.yaml

输入参数

        environment， 必填，选择一个测试环境，可选值为qa，staging等

        platform，必填，选择哪一端的代码执行，可选值为web/android/ios等，主要是mobile的seed data也在该workflow中

        instanceCount，必填，选择调用多少AWS集群机来执行本次任务

        runId，选填，填入需要重跑的test run在testrails上面的id号

        reseed_test_data，必填，选择是否执行seed data，销毁并重新生成站点测试数据

        push_report，必填，选择是否发送测试报告到99u

        image_tag，必填，选择是否获取代码库中的最新代码

jobs

restart_failed_dp:

执行./deploy/at_restart_failed_dp建立DP索引​​​

runs-on: self-hostedsteps:- name: "Checkout source code"uses: actions/checkout@v2- name: "Prepare environment"id: preprun: |source .github/global.envcat .github/global.env >> ${GITHUB_ENV}source .github/auto.envcat .github/auto.env >> ${GITHUB_ENV}if [[ "$ENVIRONMENT" == "qa" ]]; thensource .github/qa.envcat .github/qa.env >> ${GITHUB_ENV}elif [[ "$ENVIRONMENT" == "sg-staging" ]]; thensource .github/sg-staging.envcat .github/sg-staging.env >> ${GITHUB_ENV}elif [[ "$ENVIRONMENT" == "sg-prod" ]]; thensource .github/sg-prod.envcat .github/sg-prod.env >> ${GITHUB_ENV}fichmod 777 ./deploy/at_restart_failed_dp- name: AWS assume cicd roleuses: aws-actions/configure-aws-credentials@v1env:AWS_SDK_LOAD_CONFIG: falsewith:aws-region: ${{ env.AWS_REGION }}role-to-assume: ${{ env.GLOBAL_AWS_ROLE_ARN }}role-duration-seconds: 1200role-session-name: GitHubAction- name: AWS assume one-eye roleuses: aws-actions/configure-aws-credentials@v1env:AWS_SDK_LOAD_CONFIG: falsewith:aws-region: ${{ env.AWS_REGION }}role-to-assume: ${{ env.ONE_EYE_HELM_AWS_ROLE_ARN }}role-duration-seconds: 1200role-session-name: GitHubAction- name: AT Restart Failed DPid: at_restart_failed_dpuses: docker://590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/infrastructure:latestenv:AT_DP_PASSWORD: ${{ secrets.AT_DP_PASSWORD }}STAGE: ${{ env.STAGE }}TEST_AWS_REGION: ${{ env.TEST_AWS_REGION }}TEST_AWS_REGION_SHORT: ${{ env.TEST_AWS_REGION_SHORT }}KUBERNETES_NAMESPACE: ${{ env.ONE_EYE_NS }}with:entrypoint: ./deploy/at_restart_failed_dp

 备注：

• global.env存储docker注册服务地址、AWS身份认证信息、docker名称、镜像名称、主干分支名、cypress服务端口等全局参数信息
• auto.env存储AWS Region信息、K8S命名空间、AWS身份认证信息、IMAGE_TAG和自动化测试报告的S3存放url及目录信息
• qa.env存储被测系统所在的AWS Region信息、AWS身份认证信息、API接口域名信息等环境参数

seed_data:

执行./deploy/at_data_setup调用one-eye的pod造AT所需数据

runs-on: self-hostedneeds: restart_failed_dpsteps:- name: "Checkout source code"uses: actions/checkout@v2- name: "Prepare environment"id: preprun: |source .github/global.envcat .github/global.env >> ${GITHUB_ENV}source .github/auto.envcat .github/auto.env >> ${GITHUB_ENV}if [[ "$ENVIRONMENT" == "dev" ]]; thensource .github/dev.envcat .github/dev.env >> ${GITHUB_ENV}elif [[ "$ENVIRONMENT" == "nd" ]]; thensource .github/nd.envcat .github/nd.env >> ${GITHUB_ENV}elif [[ "$ENVIRONMENT" == "qa" ]]; thensource .github/qa.envcat .github/qa.env >> ${GITHUB_ENV}elif [[ "$ENVIRONMENT" == "sg-staging" ]]; thensource .github/sg-staging.envcat .github/sg-staging.env >> ${GITHUB_ENV}elif [[ "$ENVIRONMENT" == "sg-prod" ]]; thensource .github/sg-prod.envcat .github/sg-prod.env >> ${GITHUB_ENV}elseecho "Error env" $ENVIRONMENTexit 1fi- name: AWS assume cicd roleuses: aws-actions/configure-aws-credentials@v1env:AWS_SDK_LOAD_CONFIG: falsewith:aws-region: ${{ env.AWS_REGION }}role-to-assume: ${{ env.GLOBAL_AWS_ROLE_ARN }}role-duration-seconds: 1200role-session-name: GitHubAction- name: AWS assume one-eye roleuses: aws-actions/configure-aws-credentials@v1env:AWS_SDK_LOAD_CONFIG: falsewith:aws-region: ${{ env.AWS_REGION }}role-to-assume: ${{ env.ONE_EYE_HELM_AWS_ROLE_ARN }}role-duration-seconds: 1200role-session-name: GitHubAction- name: AT Data Setupid: at_data_setupuses: docker://590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/infrastructure:latestenv:RESEED_AT_DATA: ${{ env.RESEED_AT_DATA }}AT_DATA_POD_NAME: ${{ env.AT_DATA_POD_NAME }}STAGE: ${{ env.STAGE }}TEST_AWS_REGION: ${{ env.TEST_AWS_REGION }}TEST_AWS_REGION_SHORT: ${{ env.TEST_AWS_REGION_SHORT }}KUBERNETES_NAMESPACE: ${{ env.ONE_EYE_NS }}SEED_DATA_PASSWORD: ${{ env.SEED_DATA_PASSWORD }}PLATFORM: ${{ env.PLATFORM }}with:entrypoint: ./deploy/at_data_setup

execute-tests:

执行./deploy/test-execute启动K8S，分别启动initiator和worker进入测试执行

runs-on: self-hostedtimeout-minutes: 330needs: trigger_airflowsteps:- name: "Checkout source code"uses: actions/checkout@v2- name: "Prepare environment"id: preprun: |# Source our global environment variables & add them to GITHUB_ENV for subsequent stepssource .github/global.envcat .github/global.env >> ${GITHUB_ENV}# Source our environment variables for the "auto" account (where Automodo resides)source .github/auto.envcat .github/auto.env >> ${GITHUB_ENV}UNIQUE_ID=$(date +%s)echo UNIQUE_ID=$UNIQUE_ID >> $GITHUB_ENVIDENTITY_TOKEN=$(cat /var/run/secrets/eks.amazonaws.com/serviceaccount/token)echo IDENTITY_TOKEN=${IDENTITY_TOKEN} >> $GITHUB_ENV- name: AWS assume cicd roleuses: aws-actions/configure-aws-credentials@v1env:AWS_SDK_LOAD_CONFIG: falsewith:aws-region: ${{ env.AWS_REGION }}role-to-assume: ${{ env.GLOBAL_AWS_ROLE_ARN }}role-duration-seconds: 1200role-session-name: GitHubAction- name: AWS assume helm roleuses: aws-actions/configure-aws-credentials@v1env:AWS_SDK_LOAD_CONFIG: falsewith:aws-region: ${{ env.AWS_REGION }}role-to-assume: ${{ env.HELM_AWS_ROLE_ARN }}role-duration-seconds: 1200role-session-name: GitHubAction- name: Run Testsid: run-testsuses: docker://590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/infrastructure:latestenv:STAGE: autoRELEASE_NAME: ${{ env.APP_NAME }}IMAGE_NAME: ${{ env.IMAGE_NAME }}IMAGE_TAG: ${{ github.event.inputs.image_tag }}HELM_AWS_ROLE_ARN: ${{ env.HELM_AWS_ROLE_ARN }}RESEED_AT_DATA: ${{ env.RESEED_AT_DATA }}with:entrypoint: ./deploy/test-execute

Docker File

docker/Dockerfile.base

# --- Build/test ---
FROM cypress/browsers:node16.14.2-slim-chrome100-ff99-edge as base# 安装 Chrome
RUN wget -q -O - https://dl-ssl.google.com/linux/linux_signing_key.pub | apt-key add - \&& sh -c 'echo "deb [arch=amd64] http://dl.google.com/linux/chrome/deb/ stable main" >> /etc/apt/sources.list.d/google.list' \&& apt-get update \&& apt-get install -y google-chrome-stableARG ENVIRONMENT=dockerWORKDIR /# Use bash instead of sh for easier logic below
SHELL ["/bin/bash", "-c"]ADD / ./
RUN npm ci --no-optionalENV ENVIRONMENT=
ENV DIRECTOR_ADDRESS=
ENV S3_REPORT_BUCKET_URL=
ENV S3_REPORT_BUCKET=
ENV AWS_REGION=
ENV RUN_ID=
ENV CI_BUILD_ID=
ENV TYPE=CMD ENVIRONMENT=${ENVIRONMENT} AWS_REGION=${AWS_REGION} S3_REPORT_BUCKET_URL=${S3_REPORT_BUCKET_URL} S3_REPORT_BUCKET=${S3_REPORT_BUCKET} DIRECTOR_ADDRESS=${DIRECTOR_ADDRESS} RUN_ID=${RUN_ID} CI_BUILD_ID=${CI_BUILD_ID} TYPE=${TYPE} node ./cypress/scripts/cicd.js

docker/service-dockerfile.base

# --- Build/test ---
FROM cypress/browsers:node16.14.2-slim-chrome100-ff99-edge as baseWORKDIR /# Use bash instead of sh for easier logic below
SHELL ["/bin/bash", "-c"]ADD / ./RUN npm ci --no-optionalCMD node ./cypress/service/service.js

Deploy

deploy/test-execute

#!/bin/bash# echo ${IDENTITY_TOKEN} > /tmp/token
set -eunset AWS_PROFILEexport KUBECONFIG="/dev/shm/cicd"
aws --region ${AWS_REGION} eks update-kubeconfig --name=ndpg-${AWS_REGION_SHORT}-${STAGE}-eks-cluster# kubectl complains if KUBECONFIG is world readable, so lock it down
chmod 600 "${KUBECONFIG}"cd deployif [ "$RESEED_AT_DATA" = "only" ];thenecho "find flag RESEED_AT_DATA is only, Skip run tests...."exit 0
fi# Clean up the old secrets
kubectl get secrets --no-headers=true | awk '/cypress-automation-config/{print $1}'| xargs  kubectl delete secrets || true# Deploy the director if none exists
if [ "$(kubectl get service -n ${KUBERNETES_NAMESPACE} ${RELEASE_NAME}-director -o jsonpath='{.metadata.name}')" != "${RELEASE_NAME}-director" ]; thenecho -e "Starting director、dashboard、api and service...\n"helmfile -f helmfile-director.yaml apply --suppress-secrets
fi# Run the initiator job
if [ ! ${RUN_ID} ]; thenecho -e "Creating the TestRail run...\n"
elseecho -e "Rerun tests that are not passed in the existing run, ${RUN_ID}...\n"
fi#通过执行helmfile.yaml拉取镜像文件590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation:latest
INSTANCES=1 TYPE=initiator JOB_TIMEOUT=300 helmfile -f helmfile.yaml apply --suppress-secrets
#创建k8s容器如cypress-automation-qa-initiator-${UNIQUE_ID}
kubectl wait -n ${KUBERNETES_NAMESPACE} --for=condition=Complete --timeout=5m job/cypress-automation-${ENVIRONMENT}-initiator-${UNIQUE_ID}#通过执行helmfile.yaml拉取镜像文件590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation:latest
echo -e "Starting workers...\n"
TYPE=worker JOB_TIMEOUT=9000 helmfile -f helmfile.yaml apply --suppress-secrets
echo -e "See the progress at https://pangaea.testrail.io/index.php?/runs/overview/5"
#创建k8s容器如cypress-automation-qa-worker-${UNIQUE_ID}
kubectl wait -n ${KUBERNETES_NAMESPACE} --for=condition=Ready --timeout=10m -l job-name=cypress-automation-${ENVIRONMENT}-worker-${UNIQUE_ID} podsecho -e "Done."

备注：

（1）通过执行helmfile.yaml拉取镜像文件590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation:latest
（2）创建k8s容器如cypress-automation-qa-initiator-${UNIQUE_ID}
（3）通过执行helmfile.yaml拉取镜像文件590183924964.dkr.ecr.ap-southeast-1.amazonaws.com/cypress-automation:latest
（4）创建k8s容器如cypress-automation-qa-worker-${UNIQUE_ID}

当Dockerfile.base生成的镜像在容器中运行时，会执行./cypress/scripts/cicd.js开始进入测试

当service-dockerfile.base生成的镜像在容器中运行时，会执行./cypress/service/service.js启动事件监听并在监听到执行完成事件时，向99u发送报告消息

其中，initiator和worker的区别在于，initiator会在testrail中创建新的run或者获取re-run的runId