Helm 在 Cloud Build 中失败

Question

我在以下 Google 云构建步骤中使用 alpine/helm:3.0.0

- id: 'update helm app'
  name: 'alpine/helm:3.0.0'
  args: ['upgrade', 'staging', './iprocure-chart/']
  env:
  - CLOUDSDK_COMPUTE_ZONE=us-central1-a
  - CLOUDSDK_CONTAINER_CLUSTER=iprocure-cluster

问题是当我运行使用 cloud-build-local 我得到以下错误并且管道以失败结束

Starting Step #4 - "update helm app"
Step #4 - "update helm app": Already have image (with digest): alpine/helm:3.0.0
Step #4 - "update helm app": Error: UPGRADE FAILED: query: failed to query with labels: Get http://localhost:8080/api/v1/namespaces/default/secrets?labelSelector=name%3Dstaging%2Cowner%3Dhelm%2Cstatus%3Ddeployed: dial tcp 127.0.0.1:8080: connect: connection refused

Answer 1

这是因为尚未设置或通过配置。配置结帐 = https://cloud.google.com/cloud-build/docs/build-debug-locally#before_you_begin

并在您的构建步骤中添加一个 evn，如下所示：

编号：'update helm app' 姓名：'alpine/helm:3.0.0' 参数：['upgrade', 'staging', './iprocure-chart/'] 环境：
- CLOUDSDK_COMPUTE_ZONE=us-central1-a
- CLOUDSDK_CONTAINER_CLUSTER=iprocure-cluster
- KUBECONFIG=/workspace/.kube/config

如果这不起作用，请尝试在您的 helm command.Like 中传递带有 --kubeconfig 标志的配置： --kubeconfig=/workspace/.kube/config..

Helm 在 Cloud Build 中失败

Helm Fail In Cloud Build

continuous-integration

google-kubernetes-engine

kubernetes-helm

google-cloud-build