Azure Terraform(十一)Azure DevOps Pipeline 内的动态临时变量的使用
思路浅析
在我们分享的 Azure Terraform 系列文中有介绍到关于 Terraform 的状态文件远程存储的问题,我们在 Azure DevOps Pipeline 的 Task Job 加 azure_cli_script 执行内联脚本(该脚本帮我们创建好 Terraform 状态文件存储所需要的 Azure Resource Group、 Azure Storage Account、Azure KeyVault 等资源)。大家需要注意的是,内联脚本中有使用动态变量,该变量临时存储 Azure Storage Account 的 Account Key,如下图所示:
本篇文章,我继续带领大家分析如何在 Azure DevOps Pipeline 运行中创建使用动态临时变量,使用动态临时变量替换 Azure Pipeline 管道变量。
项目整体架构图
Pipeline 变量定义、输出
在此阶段,我们需要利用 azure_cli_script 任务,创建动态临时变量,输出参数,其中最主要的是将动态临时变量输出,Task yaml 如下所示
输出的变量用于同一个 stage,不同 job
- stage: script jobs: - job: azure_cli_script steps: - task: AzureCLI@2 displayName: 'Azure CLI :Create Storage Account,Key Vault And Set KeyVault Secret' name: 'output_variable' inputs: azureSubscription: 'Microsoft Azure Subscription(xxxxxxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx)' scriptType: 'bash' addSpnToEnvironment: true scriptLocation: 'inlineScript' inlineScript: | # create azure resource group az group create --location eastasia --name $(terraform_rg) # create azure storage account az storage account create --name $(storage_account) --resource-group $(terraform_rg) --location eastasia --sku Standard_LRS # create storage account container for tf state az storage container create --name $(storage_account_container) --account-name $(storage_account) # query storage key and set variable ACCOUNT_KEY=$(az storage account keys list --resource-group $(terraform_rg) --account-name $(storage_account) --query "[?keyName == 'key1'][value]" --output tsv) # create azure keyvault az keyvault create --name $(keyvault) --resource-group $(terraform_rg) --location eastasia --enable-soft-delete false # set keyvault secret,secret value is ACCOUNT_KEY az keyvault secret set --name $(keyvault_sc) --vault-name $(keyvault) --value $ACCOUNT_KEY # set secret varivale and add to environment echo "##vso[task.setvariable variable=ACCOUNT_KEY;isOutput=true]$ACCOUNT_KEY" #echo "##vso[task.setvariable variable=ACCOUNT_KEY;issecret=true;isOutput=true]$ACCOUNT_KEY" - job: same_stage_echo dependsOn: azure_cli_script variables: ACCOUNT_KEY: $[dependencies.azure_cli_script.outputs['output_variable.ACCOUNT_KEY']] steps: - task: Bash@3 displayName: 'Bash :output temporary variables in different jobs on the same stage' inputs: targetType: 'inline' script: | # echo ACCOUNT_KEY echo "ACCOUNT_KEY is $ACCOUNT_KEY"
输出变量用于不同 stage
- stage: echo_varibale dependsOn: script jobs: - job: different_stage_echo variables: ACCOUNT_KEY: $[stageDependencies.script.azure_cli_script.outputs['output_variable.ACCOUNT_KEY']] steps: - task: Bash@3 displayName: 'Bash :output temporary variables in same jobs on the same stage' inputs: targetType: 'inline' script: | # echo ACCOUNT_KEY echo "ACCOUNT_KEY is $ACCOUNT_KEY"
以下为完整的 azure-pipelines-1.yaml
# Starter pipeline # Start with a minimal pipeline that you can customize to build and deploy your code. # Add steps that build, run tests, deploy, and more: # https://aka.ms/yaml trigger: - remote_stats pool: vmImage: ubuntu-latest stages: - stage: script jobs: - job: azure_cli_script steps: - task: AzureCLI@2 displayName: 'Azure CLI :Create Storage Account,Key Vault And Set KeyVault Secret' name: 'output_variable' inputs: azureSubscription: 'Microsoft Azure Subscription(xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx)' scriptType: 'bash' addSpnToEnvironment: true scriptLocation: 'inlineScript' inlineScript: | # create azure resource group az group create --location eastasia --name $(terraform_rg) # create azure storage account az storage account create --name $(storage_account) --resource-group $(terraform_rg) --location eastasia --sku Standard_LRS # create storage account container for tf state az storage container create --name $(storage_account_container) --account-name $(storage_account) # query storage key and set variable ACCOUNT_KEY=$(az storage account keys list --resource-group $(terraform_rg) --account-name $(storage_account) --query "[?keyName == 'key1'][value]" --output tsv) # create azure keyvault az keyvault create --name $(keyvault) --resource-group $(terraform_rg) --location eastasia --enable-soft-delete false # set keyvault secret,secret value is ACCOUNT_KEY az keyvault secret set --name $(keyvault_sc) --vault-name $(keyvault) --value $ACCOUNT_KEY # set secret varivale and add to environment echo "##vso[task.setvariable variable=ACCOUNT_KEY;isOutput=true]$ACCOUNT_KEY" #echo "##vso[task.setvariable variable=ACCOUNT_KEY;issecret=true;isOutput=true]$ACCOUNT_KEY" - job: same_stage_echo dependsOn: azure_cli_script variables: ACCOUNT_KEY: $[dependencies.azure_cli_script.outputs['output_variable.ACCOUNT_KEY']] steps: - task: Bash@3 displayName: 'Bash :output temporary variables in different jobs on the same stage' inputs: targetType: 'inline' script: | # echo ACCOUNT_KEY echo "ACCOUNT_KEY is $ACCOUNT_KEY" - stage: echo_varibale dependsOn: script jobs: - job: different_stage_echo variables: ACCOUNT_KEY: $[stageDependencies.script.azure_cli_script.outputs['output_variable.ACCOUNT_KEY']] steps: - task: Bash@3 displayName: 'Bash :output temporary variables in same jobs on the same stage' inputs: targetType: 'inline' script: | # echo ACCOUNT_KEY echo "ACCOUNT_KEY is $ACCOUNT_KEY"
*****重点*****:管道内变量与动态临时变量使用区别
Pipeline 管道内使用方式:$(变量名称)
动态临时变量使用方式:$变量名称
配置 Pipeline 管道变量
使用 Azure CLI 创建 Azure Storage Account、Azure Key Vault 的内联脚本中使用管理内变量控制参数
变量名 | 变量值 |
terraform_rg | Web_Test_TF_RG |
storage_account | cnbatetfstorage |
storage_account_container | tf-state |
keyvault | cnbate-terraform-kv |
keyvault_sc | terraform-stste-storage-key |
container_key | cnbate.tf.stats |
运行 Pipeline,查看配置输出
由于我们已经在 azure-pipelines-1.yaml 文件中指定了工作分支 “remote_stats”,当我们只要触发 “remote_stats” 分支的 “push” 或者 “pull_request” 动作都会触发 Azure DevOps Pipeline 的运行。
相同 stage 内的 job 输出
不同 stage 的 job 输出
总结
本期实验,我们学习了如何在 Azure DevOps Pipeline 运行期间创建的动态临时变量以及变量的输出,使得我们更加灵活的在任意 job 中声明自定义的动态临时变量,并将动态临时变量应用到任意的 job 中,这种方式有区别与Pipeline 管道内变量,尤其是在定义阶段和使用语法上,详细内容参考官方文档。
在脚本中设置变量:https://docs.microsoft.com/en-us/azure/devops/pipelines/process/set-variables-scripts
github 代码地址:https://github.com/yunqian44/Terraform_Cnbate_Traffic_Manager
Terraform 在 Azure DevOps 中的使用系列:https://www.cnblogs.com/AllenMaster/category/1876925.html