最近对tekton, spinnaker, argo cd想系统了解一下。于是，开始有了这一个系列。一切都有套路，先从官网文档开始。

文档

https://tekton.dev/docs/getting-started/

一，tkn下载安装

这个tkn还是有点用的，比kubectl要方便快捷，功能类似，就好像istioctl这些工具一样。林子作大了，都要有自己的套装。
https://github.com/tektoncd/cli/releases
下载最新的rpm直接安装即可。
rpm -ivh tektoncd-cli-0.15.0_Linux-64bit.rpm

二，下载release.yaml

这是关键文件，下载地址：
https://storage.googleapis.com/tekton-releases/pipeline/latest/release.yaml
但这个文件，不能直接用，要修改里面的几个镜像的地址，因为都是gcr仓库，你懂。
主要修改deployment对象。
controller里的image，不少。

containers:
        - name: tekton-pipelines-controller
          image: aguncn/tektoncd-controller:v0.20.1
          args: [
            # Version, to be replace at release time
            "-version", "v0.20.1",
            # These images are built on-demand by `ko resolve` and are replaced
            # by image references by digest.
            "-kubeconfig-writer-image", "aguncn/tektoncd-kubeconfigwriter:v0.20.1", "-git-image", "aguncn/tektoncd-git-init:v0.20.1", "-entrypoint-image", "aguncn/tektoncd-entrypoint:v0.20.1", "-nop-image", "aguncn/tektoncd-nop:v0.20.1", "-imagedigest-exporter-image", "aguncn/tektoncd-imagedigestexporter:v0.20.1", "-pr-image", "aguncn/tektoncd-pullrequest-init:v0.20.1", "-build-gcs-fetcher-image", "aguncn/tektoncd-gcs-fetcher:v0.20.1",
            # This is gcr.io/google.com/cloudsdktool/cloud-sdk:302.0.0-slim
            "-gsutil-image", "aguncn/cloudsdktool-cloud-sdk",
            # The shell image must be root in order to create directories and copy files to PVCs.
            # gcr.io/distroless/base:debug as of November 15, 2020
            # image shall not contains tag, so it will be supported on a runtime like cri-o
            "-shell-image", "aguncn/distroless-base"]

webhook的一个：

containers:
        - name: webhook
          # This is the Go import path for the binary that is containerized
          # and substituted here.
          image: aguncn/tektoncd-webhook:v0.20.1
          # Resource request required for autoscaler to take any action for a metric
          resources:
            requests:
              cpu: 100m
              memory: 100Mi
            limits:
              cpu: 500m
              memory: 500Mi

至于这些aguncn/的镜像如何得到，看图呀~~
github

2021-02-03 22_01_56-悬浮球.png

dockerhub

2021-02-03 22_02_12-悬浮球.png

三，将修改过的yaml应用于k8s

kubectl apply -f release.yaml
如果没问题，两个pod都会running起来的。

2021-02-03 22_15_38-悬浮球.png

四，测试task和taskrun

task-hello.yaml

apiVersion: tekton.dev/v1beta1
kind: Task
metadata:
  name: hello-tekton
spec:
  steps:
    - name: hello-tekton
      image: rancher/library-busybox:1.31.1
      command:
        - echo
      args:
        - "Hello Tekton!"

将这个yaml apply -f 应用于集群，之后调用tkn命令启动此任务，根据提示看输出。

[root@localhost tekton]# tkn task start hello-tekton
TaskRun started: hello-tekton-run-4gf4c

In order to track the TaskRun progress run:
tkn taskrun logs hello-tekton-run-4gf4c -f -n default
[root@localhost tekton]# tkn taskrun logs hello-tekton-run-4gf4c -f -n default
[hello-tekton] Hello Tekton!

就这样吧，今天先休息。

2021-02-03 22_06_32-214 - root@localhost_~_tekton - Xshell 6 (Free for Home_School).png