kubernetes(37):持续集成(6)-部署gitlab-runner实现GitlabCI-k8s+gitlab+harbor

Gitlab CI-k8s+gitlab+harbor

https://www.qikqiak.com/k8s-book/docs/65.Gitlab%20CI.html

 

1 简介

从 Gitlab 8.0 开始,Gitlab CI 就已经集成在 Gitlab 中,我们只要在项目中添加一个.gitlab-ci.yml文件,然后添加一个Runner,即可进行持续集成。在介绍 Gitlab CI 之前,我们先看看一些 Gitlab CI 的一些相关概念。

1.1  Pipeline

可以参考Jenkins-pipeline(2):开发阶段pipeline:gitlab和Jenkins-pipeline自动集成

一次 Pipeline 其实相当于一次构建任务,里面可以包含很多个流程,如安装依赖、运行测试、编译、部署测试服务器、部署生产服务器等流程。任何提交或者 Merge Request 的合并都可以触发 Pipeline 构建,如下图所示:

+------------------+           +----------------+
|                  |  trigger  |                |
|   Commit / MR    +---------->+    Pipeline    |
|                  |           |                |
+------------------+           +----------------+

 

1.2 Stages

Stages 表示一个构建阶段,也就是上面提到的一个流程。我们可以在一次 Pipeline 中定义多个 Stages,这些 Stages 会有以下特点:

  • 所有 Stages 会按照顺序运行,即当一个 Stage 完成后,下一个 Stage 才会开始
  • 只有当所有 Stages 完成后,该构建任务 (Pipeline) 才会成功
  • 如果任何一个 Stage 失败,那么后面的 Stages 不会执行,该构建任务 (Pipeline) 失败

Stages 和 Pipeline 的关系如下所示:

|                                                        |
|  Pipeline                                              |
|                                                        |
|  +-----------+     +------------+      +------------+  |
|  |  Stage 1  |---->|   Stage 2  |----->|   Stage 3  |  |
|  +-----------+     +------------+      +------------+  |
|                                                        |
+--------------------------------------------------------+

 

 

1.3   Jobs

Jobs 表示构建工作,表示某个 Stage 里面执行的工作。我们可以在 Stages 里面定义多个 Jobs,这些 Jobs 会有以下特点:

  • 相同 Stage 中的 Jobs 会并行执行
  • 相同 Stage 中的 Jobs 都执行成功时,该 Stage 才会成功
  • 如果任何一个 Job 失败,那么该 Stage 失败,即该构建任务 (Pipeline) 失败

Jobs 和 Stage 的关系如下所示:

|                                          |

|  Stage 1                                 |

|                                          |

|  +---------+  +---------+  +---------+   |

|  |  Job 1  |  |  Job 2  |  |  Job 3  |   |

|  +---------+  +---------+  +---------+   |

|                                          |

+------------------------------------------+

 

 

 

 

2 Gitlab Runner

如果理解了上面的基本概念之后,可能我们就会发现一个问题,我们的构建任务在什么地方来执行呢,以前用 Jenkins 在 Master 和 Slave 节点都可以用来运行构建任务,而来执行我们的 Gitlab CI 构建任务的就是 Gitlab Runner。

我们知道大多数情况下构建任务都是会占用大量的系统资源的,如果直接让 Gitlab 本身来运行构建任务的话,显然 Gitlab 的性能会大幅度下降的。GitLab CI 最大的作用是管理各个项目的构建状态,因此,运行构建任务这种浪费资源的事情交给一个独立的 Gitlab Runner 来做就会好很多,更重要的是 Gitlab Runner 可以安装到不同的机器上,甚至是我们本机,这样完全就不会影响到 Gitlab 本身了。

 

安装

安装 Gitlab Runner 非常简单,我们可以完全安装官方文档:https://docs.gitlab.com/runner/install/即可,比如可以直接使用二进制、Docker 等来安装。同样的,我们这里还是将 Gitlab Runner 安装到 Kubernetes 集群中来,让我们的集群来统一管理 Gitlab 相关的服务。

2.1 验证 Kubernetes 集群

执行下面的命令验证 Kubernetes 集群:

# kubectl cluster-info
Kubernetes master is running at https://10.6.76.25:6443
Alertmanager is running at https://10.6.76.25:6443/api/v1/namespaces/kube-system/services/alertmanager:http/proxy
KubeDNS is running at https://10.6.76.25:6443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy
kube-state-metrics is running at https://10.6.76.25:6443/api/v1/namespaces/kube-system/services/kube-state-metrics:http-metrics/proxy
NodeExporter is running at https://10.6.76.25:6443/api/v1/namespaces/kube-system/services/node-exporter:metrics/proxy
Prometheus is running at https://10.6.76.25:6443/api/v1/namespaces/kube-system/services/prometheus:http/proxy

To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'.

#cluster-info这个命令会显示当前链接的集群状态和可用的集群服务列表。

 

 

2.2 获取 Gitlab CI Register Token

kubernetes(36):持续集成(5)-k8s集群中搭建gitlab

前面我们已经成功安装了 Gitlab,在浏览器中打开gitlab.xxx.com页面,然后登录后进入到管理页面http://gitlab.xxx.com/admin,然后点击导航栏中的Runner,可以看到该页面中有两个总要的参数,一个是 URL,另外一个就是 Register Token,下面的步骤中需要用到这两个参数值。

 

 

注意:不要随便泄露 Token

 

 

2.3 编写 Gitlab CI Runner 资源清单文件

runner-cm.yaml

同样将 Runner 相关的资源对象都安装到kube-ops这个 namespace 下面,首先,我们通过 ConfigMap 资源来传递 Runner 镜像所需的环境变量(runner-cm.yaml):

apiVersion: v1
data:
  REGISTER_NON_INTERACTIVE: "true"
  REGISTER_LOCKED: "false"
  METRICS_SERVER: "0.0.0.0:9100"
  CI_SERVER_URL: "http://gitlab.kube-ops.svc.cluster.local/ci"
  RUNNER_REQUEST_CONCURRENCY: "4"
  RUNNER_EXECUTOR: "kubernetes"
  KUBERNETES_NAMESPACE: "kube-ops"
  KUBERNETES_PRIVILEGED: "true"
  KUBERNETES_CPU_LIMIT: "1"
  KUBERNETES_CPU_REQUEST: "500m"
  KUBERNETES_MEMORY_LIMIT: "1Gi"
  KUBERNETES_SERVICE_CPU_LIMIT: "1"
  KUBERNETES_SERVICE_MEMORY_LIMIT: "1Gi"
  KUBERNETES_HELPER_CPU_LIMIT: "500m"
  KUBERNETES_HELPER_MEMORY_LIMIT: "100Mi"
  KUBERNETES_PULL_POLICY: "if-not-present"
  KUBERNETES_TERMINATIONGRACEPERIODSECONDS: "10"
  KUBERNETES_POLL_INTERVAL: "5"
  KUBERNETES_POLL_TIMEOUT: "360"
kind: ConfigMap
metadata:
  labels:
    app: gitlab-ci-runner
  name: gitlab-ci-runner-cm
  namespace: kube-ops

 

要注意CI_SERVER_URL对应的值需要指向我们的 Gitlab 实例的 URL(可以是外网地址,也可以是 Kubernetes 集群内部的 Service DNS 地址,因为 Runner 也是运行在 Kubernetes 集群中的),并加上/cihttp://gitlab.kube-ops.svc.cluster.local/ci )。此外还添加了一些构建容器运行的资源限制,可以自己根据需要进行更改即可。

注意:在向 ConfigMap 添加新选项后,需要删除 GitLab CI Runner Pod。因为我们是使用 envFrom来注入上面的这些环境变量而不是直接使用env的(envFrom 通过将环境变量放置到ConfigMapsSecrets来帮助减小清单文件。

 

 

 

另外如果要添加其他选项的话,我们可以在 Pod 中运行gitlab-ci-multi-runner register --help命令来查看所有可使用的选项,只需为要配置的标志添加 env 变量即可,如下所示:

gitlab-runner@gitlab-ci-runner-0:/$ gitlab-ci-multi-runner register --help
[...]
--kubernetes-cpu-limit value                          The CPU allocation given to build containers (default: "1") [$KUBERNETES_CPU_LIMIT]
--kubernetes-memory-limit value                       The amount of memory allocated to build containers (default: "4Gi") [$KUBERNETES_MEMORY_LIMIT]
--kubernetes-service-cpu-limit value                  The CPU allocation given to build service containers (default: "1") [$KUBERNETES_SERVICE_CPU_LIMIT]
--kubernetes-service-memory-limit value               The amount of memory allocated to build service containers (default: "1Gi") [$KUBERNETES_SERVICE_MEMORY_LIMIT]
--kubernetes-helper-cpu-limit value                   The CPU allocation given to build helper containers (default: "500m") [$KUBERNETES_HELPER_CPU_LIMIT]
--kubernetes-helper-memory-limit value                The amount of memory allocated to build helper containers (default: "3Gi") [$KUBERNETES_HELPER_MEMORY_LIMIT]
--kubernetes-cpu-request value                        The CPU allocation requested for build containers [$KUBERNETES_CPU_REQUEST]
...
--pre-clone-script value                              Runner-specific command script executed before code is pulled [$RUNNER_PRE_CLONE_SCRIPT]
[...]

 

 

如果定义的 Gitlab 域名并不是通过外网的 DNS 进行解析的,而是通过 /etc/hosts 俩进行映射的,那么我们就需要在 runner 的 Pod 中去添加 git.qikqiak.com 对应的 hosts 了,那么如何添加呢?我们可以想到的是 Pod 的 hostAlias 可以实现这个需求,但是 runner 的 Pod 是自动生成的,没办法直接去定义 hostAlias。这里我们就可以通过上面的--pre-clone-script参数来指定一段脚本来添加 hosts 信息,也就是在上面的 ConfigMap 中添加环境变量RUNNER_PRE_CLONE_SCRIPT的值即可:

RUNNER_PRE_CLONE_SCRIPT = "echo 'xx.xx.xxx.xx git.qikqiak.com' >> /etc/hosts"
#这个后面就用到了

runner-scripts-cm.yaml

除了上面的一些环境变量相关的配置外,还需要一个用于注册、运行和取消注册 Gitlab CI Runner 的小脚本。只有当 Pod 正常通过 Kubernetes(TERM信号)终止时,才会触发转轮取消注册。 如果强制终止 Pod(SIGKILL信号),Runner 将不会注销自身。必须手动完成对这种死的 Runner 的清理,配置清单文件如下:(runner-scripts-cm.yaml)

 

apiVersion: v1
data:
  run.sh: |
    #!/bin/bash
    unregister() {
        kill %1
        echo "Unregistering runner ${RUNNER_NAME} ..."
        /usr/bin/gitlab-ci-multi-runner unregister -t "$(/usr/bin/gitlab-ci-multi-runner list 2>&1 | tail -n1 | awk '{print $4}' | cut -d'=' -f2)" -n ${RUNNER_NAME}
        exit $?
    }
    trap 'unregister' EXIT HUP INT QUIT PIPE TERM
    echo "Registering runner ${RUNNER_NAME} ..."
    /usr/bin/gitlab-ci-multi-runner register -r ${GITLAB_CI_TOKEN}
    sed -i 's/^concurrent.*/concurrent = '"${RUNNER_REQUEST_CONCURRENCY}"'/' /home/gitlab-runner/.gitlab-runner/config.toml
    echo "Starting runner ${RUNNER_NAME} ..."
    /usr/bin/gitlab-ci-multi-runner run -n ${RUNNER_NAME} &
    wait
kind: ConfigMap
metadata:
  labels:
    app: gitlab-ci-runner
  name: gitlab-ci-runner-scripts
  namespace: kube-ops

 

 

 

 

 

可以看到需要一个 GITLAB_CI_TOKEN,然后我们用 Gitlab CI runner token 来创建一个 Kubernetes secret 对象。将 token 进行 base64 编码:

 

 

 

# echo "vjrZr36yypzdXnBSj6sy" |  base64 -w0
dmpyWnIzNnl5cHpkWG5CU2o2c3kK

 

gitlab-ci-token-secret.yaml

然后使用上面的 token 创建一个 Secret 对象:(gitlab-ci-token-secret.yaml)

 

apiVersion: v1
kind: Secret
metadata:
  name: gitlab-ci-token
  namespace: kube-ops
  labels:
    app: gitlab-ci-runner
data:
  GITLAB_CI_TOKEN: dmpyWnIzNnl5cHpkWG5CU2o2c3kK

 

 

# kubectl apply -f gitlab-ci-token-secret.yaml
secret/gitlab-ci-token created

 

 

runner-statefulset.yaml

然后接下来就可以来编写一个用于真正运行 Runner 的控制器对象,我们这里使用 Statefulset。首先,在开始运行的时候,尝试取消注册所有的同名 Runner,当节点丢失时(即NodeLost事件),这尤其有用。然后再尝试重新注册自己并开始运行。在正常停止 Pod 的时候,Runner 将会运行unregister命令来尝试取消自己,所以 Gitlab 就不能再使用这个 Runner 了,这个是通过 Kubernetes Pod 生命周期中的hooks来完成的。

另外通过使用envFrom来指定SecretsConfigMaps来用作环境变量,对应的资源清单文件如下:

(runner-statefulset.yaml)

 

apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
  name: gitlab-ci-runner
  namespace: kube-ops
  labels:
    app: gitlab-ci-runner
spec:
  updateStrategy:
    type: RollingUpdate
  replicas: 2
  serviceName: gitlab-ci-runner
  template:
    metadata:
      labels:
        app: gitlab-ci-runner
    spec:
      volumes:
      - name: gitlab-ci-runner-scripts
        projected:
          sources:
          - configMap:
              name: gitlab-ci-runner-scripts
              items:
              - key: run.sh
                path: run.sh
                mode: 0755
      serviceAccountName: gitlab-ci
      securityContext:
        runAsNonRoot: true
        runAsUser: 999
        supplementalGroups: [999]
      containers:
      - image: gitlab/gitlab-runner:latest
        name: gitlab-ci-runner
        command:
        - /scripts/run.sh
        envFrom:
        - configMapRef:
            name: gitlab-ci-runner-cm
        - secretRef:
            name: gitlab-ci-token
        env:
        - name: RUNNER_NAME
          valueFrom:
            fieldRef:
              fieldPath: metadata.name
        ports:
        - containerPort: 9100
          name: http-metrics
          protocol: TCP
        volumeMounts:
        - name: gitlab-ci-runner-scripts
          mountPath: "/scripts"
          readOnly: true
      restartPolicy: Always

 

runner-rbac.yaml

可以看到上面我们使用了一个名为 gitlab-ci 的 serviceAccount,新建一个 rbac 资源清单文件:(runner-rbac.yaml)

 

apiVersion: v1
kind: ServiceAccount
metadata:
  name: gitlab-ci
  namespace: kube-ops
---
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: gitlab-ci
  namespace: kube-ops
rules:
  - apiGroups: [""]
    resources: ["*"]
    verbs: ["*"]
---
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: gitlab-ci
  namespace: kube-ops
subjects:
  - kind: ServiceAccount
    name: gitlab-ci
    namespace: kube-ops
roleRef:
  kind: Role
  name: gitlab-ci
  apiGroup: rbac.authorization.k8s.io

 

 

2.4    创建 Runner 资源对象

# ls gitlab-ci-token-secret.yaml runner-*
gitlab-ci-token-secret.yaml  runner-cm.yaml  runner-rbac.yaml  runner-statefulset.yaml  runner-scripts-cm.yaml

kubectl apply -f runner-cm.yaml
kubectl apply -f runner-scripts-cm.yaml
kubectl apply -f gitlab-ci-token-secret.yaml
kubectl apply -f runner-rbac.yaml
kubectl apply -f runner-statefulset.yaml

 

创建完成后,可以通过查看 Pod 状态判断 Runner 是否运行成功:

 

# kubectl -n kube-ops  get pod| grep gitlab-ci
gitlab-ci-runner-0                             1/1     Running   0          4m35s
gitlab-ci-runner-1                             1/1     Running   0          116s

 

以看到已经成功运行了两个(具体取决于StatefulSet清单中的副本数) Runner 实例,然后切换到 Gitlab Admin 页面下面的 Runner 页面:

 

 

当然我们也可以根据需要更改 Runner 的一些配置,比如添加 tag 标签等。

 

 

3 Gitlab CI

准备测试代码

下载代码上传到gitlab

测试项cnych/presentation-gitlab-k8s 我fork了一下

接下来使用 Gitlab CI 所用到的代码库可以从 Github 上获得:https://github.com/wangxu01/gitlab-ci-k8s-demo ,可以在 Gitlab 上新建一个项目导入该仓库,当然也可以新建一个空白的仓库,然后将 Github 上面的项目 Clone 到本地后,更改远程仓库地址即可:

 

[root@k8s-master test]# git clone https://github.com/wangxu01/gitlab-ci-k8s-demo.git
正克隆到 'gitlab-ci-k8s-demo'...
remote: Enumerating objects: 2036, done.
remote: Total 2036 (delta 0), reused 0 (delta 0), pack-reused 2036
接收对象中: 100% (2036/2036), 3.29 MiB | 767.00 KiB/s, done.
处理 delta 中: 100% (787/787), done.
[root@k8s-master test]# cd gitlab-ci-k8s-demo/
[root@k8s-master gitlab-ci-k8s-demo]# ls
Dockerfile  Gopkg.lock  Gopkg.toml  LICENSE  main.go  Makefile  manifests  README.md  vendor  VERSION
[root@k8s-master gitlab-ci-k8s-demo]# git remote set-url origin ssh://git@gitlab.wangxu.com:30022/root/presentation-gitlab-k8s.git
[root@k8s-master gitlab-ci-k8s-demo]# git push -u origin master
Counting objects: 2036, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (1133/1133), done.
Writing objects: 100% (2036/2036), 3.29 MiB | 0 bytes/s, done.
Total 2036 (delta 787), reused 2036 (delta 787)
remote: Resolving deltas: 100% (787/787), done.
remote:
remote: The private project root/presentation-gitlab-k8s was successfully created.
remote:
remote: To configure the remote, run:
remote:   git remote add origin git@gitlab.wangxu.com:root/presentation-gitlab-k8s.git
remote:
remote: To view the project, visit:
remote:   http://gitlab.wangxu.com/root/presentation-gitlab-k8s
remote:
To ssh://git@gitlab.wangxu.com:30022/root/presentation-gitlab-k8s.git
 * [new branch]      master -> master
分支 master 设置为跟踪来自 origin 的远程分支 master。
[root@k8s-master gitlab-ci-k8s-demo]#

 

自动触发部署

当我们把仓库推送到 Gitlab 以后,应该可以看到 Gitlab CI 开始执行构建任务了:

 

 

 

 

 

此时 Runner Pod 所在的 namespace 下面也会出现两个新的 Pod:

 

# kubectl get pods -n kube-ops
……
runner-aczwwb3k-project-2-concurrent-0chkmk    0/2     ContainerCreating   0          2m14s
runner-aczwwb3k-project-2-concurrent-1nbfbv    0/2     ContainerCreating   0          2m13s

 

 

错误1-gitlab域名

这两个新的 Pod 就是用来执行具体的 Job 任务的,这里同时出现两个证明第一步是并行执行的两个任务,从上面的 Pipeline 中也可以看到是 test 和 test2 这两个 Job。我们可以看到在执行 image_build 任务的时候出现了错误:

 

 

 

可以点击查看这个 Job 失败详细信息:

Running with gitlab-runner 12.4.0 (1564076b)
  on gitlab-ci-runner-1 DQisyaVT
Using Kubernetes namespace: kube-ops
Using Kubernetes executor with image golang:1.10.3-stretch ...
Waiting for pod kube-ops/runner-dqisyavt-project-2-concurrent-07lr98 to be running, status is Pending
Running on runner-dqisyavt-project-2-concurrent-07lr98 via gitlab-ci-runner-1...
Fetching changes with git depth set to 50...
Initialized empty Git repository in /builds/root/presentation-gitlab-k8s/.git/
Created fresh repository.
fatal: unable to access 'http://gitlab-ci-token:[MASKED]@gitlab.wangxu.com/root/presentation-gitlab-k8s.git/': Could not resolve host: gitlab.wangxu.com
ERROR: Job failed: command terminated with exit code 1

 


 

这是因为我的gitlab域名没有在 run pod进行解析

# cat runner-cm.yaml
apiVersion: v1
data:
  REGISTER_NON_INTERACTIVE: "true"
  REGISTER_LOCKED: "false"
  METRICS_SERVER: "0.0.0.0:9100"
  CI_SERVER_URL: "http://gitlab.kube-ops.svc.cluster.local/ci"
  RUNNER_REQUEST_CONCURRENCY: "4"
  RUNNER_EXECUTOR: "kubernetes"
  KUBERNETES_NAMESPACE: "kube-ops"
  KUBERNETES_PRIVILEGED: "true"
  KUBERNETES_CPU_LIMIT: "1"
  KUBERNETES_CPU_REQUEST: "500m"
  KUBERNETES_MEMORY_LIMIT: "1Gi"
  KUBERNETES_SERVICE_CPU_LIMIT: "1"
  KUBERNETES_SERVICE_MEMORY_LIMIT: "1Gi"
  KUBERNETES_HELPER_CPU_LIMIT: "500m"
  KUBERNETES_HELPER_MEMORY_LIMIT: "100Mi"
  KUBERNETES_PULL_POLICY: "if-not-present"
  KUBERNETES_TERMINATIONGRACEPERIODSECONDS: "10"
  KUBERNETES_POLL_INTERVAL: "5"
  KUBERNETES_POLL_TIMEOUT: "360"
  RUNNER_PRE_CLONE_SCRIPT : "echo '10.6.76.23 gitlab.wangxu.com' >> /etc/hosts"
  RUNNER_PRE_CLONE_SCRIPT : "echo '10.6.76.24 gitlab.wangxu.com' >> /etc/hosts"
kind: ConfigMap
metadata:
  labels:
    app: gitlab-ci-runner
  name: gitlab-ci-runner-cm
  namespace: kube-ops
#重启runner
kubectl apply -f  runner-cm.yaml
kubectl delete -f  runner-statefulset.yaml
kubectl apply -f  runner-statefulset.yaml

 

 

错误2-代码配置文件

 

下载代码通过了,但在执行 image_build 任务的时候出现了错误:

WARNING! Using --password via the CLI is insecure. Use --password-stdin.
Error response from daemon: Get http://registry.qikqiak.com/v1/users/: dial tcp: lookup registry.qikqiak.com on 10.96.0.10:53: no such host

 

 

这里调用的还是原git的地址

 

我们把原项目的registry.qikqiak.com 改成自己的gitlab.wangxu.com 再提交一下

 

[root@k8s-master gitlab-ci-k8s-demo]# grep -r   registry.qikqiak.com  *
manifests/deployment.yaml:        image: registry.qikqiak.com/gitdemo/gitlab-k8s:__VERSION__
[root@k8s-master gitlab-ci-k8s-demo]# vim manifests/deployment.yaml
[root@k8s-master gitlab-ci-k8s-demo]# grep -r   image:  *
manifests/deployment.yaml:        image: harbor.wangxu.com/gitdemo/gitlab-k8s:__VERSION__
vendor/golang.org/x/crypto/openpgp/packet/userattribute_test.go:                        t.Errorf("Error decoding JPEG image:%v", err)
[root@k8s-master gitlab-ci-k8s-demo]# git add .
[root@k8s-master gitlab-ci-k8s-demo]# git commit -m 'change image'
[master 532d6d0] change image
 1 file changed, 1 insertion(+), 1 deletion(-)
(reverse-i-search)`': ^C
[root@k8s-master gitlab-ci-k8s-demo]# git push origin master
Counting objects: 7, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (4/4), done.
Writing objects: 100% (4/4), 357 bytes | 0 bytes/s, done.
Total 4 (delta 3), reused 0 (delta 0)
To ssh://git@gitlab.wangxu.com:30022/root/presentation-gitlab-k8s.git
   3fec16d..532d6d0  master -> master
[root@k8s-master gitlab-ci-k8s-demo]#

 

我们也没有在 Gitlab 中开启 Container Registry,所以环境变量中并没有这些值,我们这里使用 Harbor 来作为我们的镜像仓库,这里我们只需要把 Harbor 相关的配置以参数的形式配置到环境中就可以了。

定位到项目 -> 设置 -> CI/CD,展开Environment variables栏目,配置镜像仓库相关的参数值:

 

 

 

 

 

 

 

配置上后,我们在上面失败的 Job 任务上点击“重试”,在重试过后依然可以看到会出现下面的错误信息:

 

$ docker login -u "${CI_REGISTRY_USER}" -p "${CI_REGISTRY_PASSWORD}" registry.qikqiak.com
WARNING! Using --password via the CLI is insecure. Use --password-stdin.
Error response from daemon: Get http://registry.qikqiak.com/v1/users/: dial tcp: lookup registry.qikqiak.com on 10.96.0.10:53: no such host
ERROR: Job failed: command terminated with exit code 1

 

原地址还是没有改过来,这个其实是配置在.gitlab-ci.yml文件

 

[root@k8s-master gitlab-ci-k8s-demo]# sed -i 's/git.qikqiak.com/gitlab.wangxu.com/g' .gitlab-ci.yml
[root@k8s-master gitlab-ci-k8s-demo]# sed -i 's/registry.qikqiak.com/harbor.wangxu.com/g' .gitlab-ci.yml
[root@k8s-master gitlab-ci-k8s-demo]# git add .
[root@k8s-master gitlab-ci-k8s-demo]# git commit -m 'change 5'
[master dce9e4c] change 5
 1 file changed, 5 insertions(+), 5 deletions(-)
[root@k8s-master gitlab-ci-k8s-demo]# git push origin master
Counting objects: 5, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (3/3), done.
Writing objects: 100% (3/3), 319 bytes | 0 bytes/s, done.
Total 3 (delta 2), reused 0 (delta 0)
To ssh://git@gitlab.wangxu.com:30022/root/presentation-gitlab-k8s.git
   acbed8f..dce9e4c  master -> master
[root@k8s-master gitlab-ci-k8s-demo]#

 

 

错误3-harbor域名

 

$ docker login -u "${CI_REGISTRY_USER}" -p "${CI_REGISTRY_PASSWORD}" harbor.wangxu.com
WARNING! Using --password via the CLI is insecure. Use --password-stdin.
Error response from daemon: Get http://harbor.wangxu.com/v1/users/: dial tcp: lookup harbor.wangxu.com on 10.96.0.10:53: no such host

 

 

还是域名的问题

#vim runner-cm.yaml
...
  RUNNER_PRE_CLONE_SCRIPT : "echo '10.6.76.23 gitlab.wangxu.com harbor.wangxu.com' >> /etc/hosts"
  RUNNER_PRE_CLONE_SCRIPT : "echo '10.6.76.24 gitlab.wangxu.com harbor.wangxu.com' >> /etc/hosts"
...

 

 

kubectl apply -f  runner-cm.yaml
kubectl delete -f  runner-statefulset.yaml
kubectl apply -f  runner-statefulset.yaml

 

 

错误4-配置环境变量

然后又失败了

$ docker login -u "${CI_REGISTRY_USER}" -p "${CI_REGISTRY_PASSWORD}" harbor.wangxu.com
WARNING! Using --password via the CLI is insecure. Use --password-stdin.
Error response from daemon: Get https://harbor.wangxu.com/v2/: unauthorized: authentication required
ERROR: Job failed: command terminated with exit code 1

 

Harbor密码写错了 选中受保护

 

 

 

 

编译完成

终于过去了

 

 

Running with gitlab-runner 12.4.0 (1564076b)
  on gitlab-ci-runner-0 ztHmfWd9
Using Kubernetes namespace: kube-ops
Using Kubernetes executor with image docker:latest ...
Waiting for pod kube-ops/runner-zthmfwd9-project-3-concurrent-06mlgk to be running, status is Pending
Running on runner-zthmfwd9-project-3-concurrent-06mlgk via gitlab-ci-runner-0...
$ echo '10.6.76.24 gitlab.wangxu.com harbor.wangxu.com' >> /etc/hosts
Fetching changes with git depth set to 50...
Initialized empty Git repository in /builds/root/presentation-gitlab-k8s/.git/
Created fresh repository.
From http://gitlab.wangxu.com/root/presentation-gitlab-k8s
 * [new branch]      master     -> origin/master
Checking out d4bfe70b as master...

Skipping Git submodules setup
Downloading artifacts for compile (126)...
Downloading artifacts from coordinator... ok        id=126 responseStatus=200 OK token=4yKS-1uT
$ mkdir -p "/go/src/gitlab.wangxu.com/${CI_PROJECT_NAMESPACE}"
$ ln -sf "${CI_PROJECT_DIR}" "/go/src/gitlab.wangxu.com/${CI_PROJECT_PATH}"
$ cd "/go/src/gitlab.wangxu.com/${CI_PROJECT_PATH}/"
$ docker info
Client:
 Debug Mode: false

Server:
 Containers: 0
  Running: 0
  Paused: 0
  Stopped: 0
 Images: 0
 Server Version: 17.03.2-ce
 Storage Driver: overlay
  Backing Filesystem: xfs
  Supports d_type: false
 Logging Driver: json-file
 Cgroup Driver: cgroupfs
 Plugins:
  Volume: local
  Network: bridge host macvlan null overlay
  Log: 
 Swarm: inactive
 Runtimes: runc
 Default Runtime: runc
 Init Binary: docker-init
 containerd version: 4ab9917febca54791c5f071a9d1f404867857fcc
 runc version: 54296cf40ad8143b62dbcaa1d90e520a2136ddfe
 init version: 949e6fa
 Security Options:
  seccomp
   Profile: default
 Kernel Version: 5.2.10-1.el7.elrepo.x86_64
 Operating System: Alpine Linux v3.5 (containerized)
 OSType: linux
 Architecture: x86_64
 CPUs: 4
 Total Memory: 7.783GiB
 Name: runner-zthmfwd9-project-3-concurrent-06mlgk
 ID: S4QT:6C5I:BKXI:3HX3:FTRQ:ZGO2:2AUP:X5G5:YR2I:MHUV:XMYE:XTWU
 Docker Root Dir: /var/lib/docker
 Debug Mode: false
 Registry: https://index.docker.io/v1/
 Experimental: false
 Insecure Registries:
  harbor.wangxu.com
  127.0.0.0/8
 Live Restore Enabled: false

WARNING: overlay: the backing xfs filesystem is formatted without d_type support, which leads to incorrect behavior.
         Reformat the filesystem with ftype=1 to enable d_type support.
         Running without d_type support will not be supported in future releases.
WARNING: bridge-nf-call-iptables is disabled
WARNING: bridge-nf-call-ip6tables is disabled
$ docker login -u "${CI_REGISTRY_USER}" -p "${CI_REGISTRY_PASSWORD}" "${CI_REGISTRY}"
WARNING! Using --password via the CLI is insecure. Use --password-stdin.
WARNING! Your password will be stored unencrypted in /root/.docker/config.json.
Configure a credential helper to remove this warning. See
https://docs.docker.com/engine/reference/commandline/login/#credentials-store

Login Succeeded
$ docker build -t "${CI_REGISTRY_IMAGE}:latest" .
Sending build context to Docker daemon  28.63MB

Step 1/4 : FROM busybox:1.28.4-glibc
1.28.4-glibc: Pulling from library/busybox
5497949500d2: Pulling fs layer
5497949500d2: Verifying Checksum
5497949500d2: Download complete
5497949500d2: Pull complete
Digest: sha256:bda689514be526d9557ad442312e5d541757c453c50b8cf2ae68597c291385a1
Status: Downloaded newer image for busybox:1.28.4-glibc
 ---> 56cc512116c8
Step 2/4 : COPY app /bin/app
 ---> b97b889363e3
Removing intermediate container 197636a53e2a
Step 3/4 : RUN chmod +x /bin/app
 ---> Running in 0a69653107b2
 ---> d1ee3ebf5944
Removing intermediate container 0a69653107b2
Step 4/4 : CMD /bin/app
 ---> Running in 74ff2064d7c8
 ---> a3f04c58f4bd
Removing intermediate container 74ff2064d7c8
Successfully built a3f04c58f4bd
$ docker tag "${CI_REGISTRY_IMAGE}:latest" "${CI_REGISTRY}/library/${CI_REGISTRY_IMAGE}:${CI_COMMIT_REF_NAME}"
$ test ! -z "${CI_COMMIT_TAG}" && docker push "${CI_REGISTRY_IMAGE}:latest"
$ docker push "${CI_REGISTRY}/library/${CI_REGISTRY_IMAGE}:${CI_COMMIT_REF_NAME}"
The push refers to a repository [harbor.wangxu.com/library/gitlab-ci-k8s-demo]
b3ca74fc4d32: Preparing
e49dd1e534d9: Preparing
e49dd1e534d9: Pushed
b3ca74fc4d32: Pushed
master: digest: sha256:37e5a5d0935f69926e9fa66bc7b7488272c07cdbd7d26666f38e4ef24bd209dc size: 739
Job succeeded

可以看到镜像传到harbor上了

 

 

 

 

配置关联的k8s集群

然后我们又失败了

 

 

 

       在最后的阶段deploy_review可以看到失败了,这是因为在最后的部署阶段我们使用kubectl工具操作集群的时候并没有关联上任何集群。

我们在 Gitlab CI 中部署阶段使用到的镜像是cnych/kubectl,该镜像的Dockerfile文件可以在仓库https://github.com/cnych/docker-kubectl中获取

我们也fork一下  https://github.com/wangxu01/docker-kubectl

cnych/kubectl 我们下载  再上传一个 dockerHub wangxu01/kubelet

谢谢作者

 

#kubelet Dockerfile

FROM alpine:3.8

MAINTAINER cnych <icnych@gmail.com>

ENV KUBE_LATEST_VERSION="v1.13.4"

RUN apk add --update ca-certificates \
 && apk add --update -t deps curl \
 && apk add --update gettext \
 && apk add --update git \
 && curl -L https://storage.googleapis.com/kubernetes-release/release/${KUBE_LATEST_VERSION}/bin/linux/amd64/kubectl -o /usr/local/bin/kubectl \
 && chmod +x /usr/local/bin/kubectl \
 && apk del --purge deps \
 && rm /var/cache/apk/*

 ENTRYPOINT ["kubectl"]
 CMD ["--help"]

 

 

我们知道kubectl在使用的时候默认会读取当前用户目录下面的~/.kube/config文件来链接集群,当然我们可以把连接集群的信息直接内置到上面的这个镜像中去,这样就可以直接操作集群了,但是也有一个不好的地方就是不方便操作,假如要切换一个集群还得重新制作一个镜像。所以一般我们这里直接在 Gitlab 上配置集成 Kubernetes 集群。

 

 

在项目页面点击Add Kubernetes Cluster -> Add existing cluster

 

 

 

1.Kubernetes cluster name 可以随便填

2.API URL 是你的集群的apiserver的地址, 一般可以通过输入kubectl cluster-info获取,Kubernetes master 地址就是需要的

# kubectl cluster-info | grep 'Kubernetes master' | awk '/http/ {print $NF}'
https://10.6.76.25:6443

 

3.CA证书、Token、项目命名空间

对于我们这个项目准备部署在一个名为gitlab的 namespace 下面,所以首先我们需要到目标集群中创建一个 namespace:

kubectl create ns gitlab

 

由于我们在部署阶段需要去创建、删除一些资源对象,所以我们也需要对象的 RBAC 权限,这里为了简单,我们直接新建一个 ServiceAccount,绑定上一个cluster-admin的权限:(gitlab-sa.yaml)

 

---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: gitlab
  namespace: gitlab

---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: ClusterRoleBinding
metadata:
  name: gitlab
  namespace: gitlab
subjects:
  - kind: ServiceAccount
    name: gitlab
    namespace: gitlab
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin

 

然后创建上面的 ServiceAccount 对象:

$ kubectl apply -f sa.yaml
serviceaccount "gitlab" created
clusterrolebinding.rbac.authorization.k8s.io "gitlab" created

 

可以通过上面创建的 ServiceAccount 获取 CA 证书和 Token:

 

$ kubectl get serviceaccount gitlab -n gitlab -o json | jq -r '.secrets[0].name'
gitlab-token-f9zp7

然后根据上面的Secret找到CA证书

$ kubectl get secret gitlab-token-f9zp7 -n gitlab -o json | jq -r '.data["ca.crt"]' | base64 -d
xxxxxCA证书内容xxxxx

 

当然要找到对应的 Token 也很简单

$ kubectl get secret gitlab-token-f9zp7  -n gitlab -o json | jq -r '.data.token' | base64 -d
xxxxxxtoken值xxxx
[root@k8s-master gitlab]# kubectl get serviceaccount gitlab -n gitlab -o json | jq -r '.secrets[0].name'
gitlab-token-lx7kn
[root@k8s-master gitlab]# kubectl get secret gitlab-token-lx7kn -n gitlab -o json | jq -r '.data["ca.crt"]' | base64 -d
-----BEGIN CERTIFICATE-----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-----END CERTIFICATE-----
[root@k8s-master gitlab]# kubectl get secret  gitlab-token-lx7kn -n gitlab -o json | jq -r '.data.token' | base64 -d
eyJhbGciOiJSUzI1NiIsImtpZCI6IiJ9.eyJpc3MiOiJrdWJlcm5ldGVzL3NlcnZpY2VhY2NvdW50Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9uYW1lc3BhY2UiOiJnaXRsYWIiLCJrdWJlcm5ldGVzLmlvL3NlcnZpY2VhY2NvdW50L3NlY3JldC5uYW1lIjoiZ2l0bGFiLXRva2VuLWx4N2tuIiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZXJ2aWNlLWFjY291bnQubmFtZSI6ImdpdGxhYiIsImt1YmVybmV0ZXMuaW8vc2VydmljZWFjY291bnQvc2VydmljZS1hY2NvdW50LnVpZCI6IjFhYWRjNzljLTQ2MGYtNDMwNy05MmI1LTNkOTgwMmIyOTExZiIsInN1YiI6InN5c3RlbTpzZXJ2aWNlYWNjb3VudDpnaXRsYWI6Z2l0bGFiIn0.04Kfa1xY0_g4xUE_CqMNvtAzJGlO7CICv1iicP7WzMiTEjKa6c031duYBP85-jeRdrHxnEX2VaUrqJIS5eT-GJajBMrUD6D55qa_N5OBrUdTwxe-NJZigYKL3xE1fpZKKoJg5ggGd239BJduUvp10Xo_V0hhwhgsGmHDzMyhtZEzWmty3DjiNFBNedxQwV2fNoOy0Z8UYq-A38hGpQl_zPt_TVdsMSgw2X8KIvgWetsOspfAl6DL0URUmjBSXhKqEl1g4YNGin-cZGyDWeUbjg3LpoXQzediyN5tmI2pwmW-tgRZqnK7vLZQQ1zQ7fmD27l8orMNmMhQmt6zVdLJng

 

 

 

is blocked: Requests to the local network are not allowed

https://blog.csdn.net/weixin_37417954/article/details/83106881

 

 

 

接着就保存成功了

 

 

 

 

 

 

 

.gitlab-ci.yml

现在 Gitlab CI 的环境都准备好了,我们可以来看下用于描述 Gitlab CI 的.gitlab-ci.yml文件。

一个 Job 在.gitlab-ci.yml文件中一般如下定义:

# 运行golang测试用例
test:
  stage: test
  script:
    - go test ./...

 

上面这个 Job 会在 test 这个 Stage 阶段运行。

为了指定运行的 Stage 阶段,可以在.gitlab-ci.yml文件中放置任意一个简单的列表:

 

# 所有 Stage
stages:
  - test
  - build
  - release
  - deploy

 

 

你可以指定用于在全局或者每个作业上执行命令的镜像:

# 对于未指定镜像的作业,会使用下面的镜像
image: golang:1.10.3-stretch
# 或者对于特定的job使用指定的镜像
test:
  stage: test
  image: python:3
  script:
    - echo Something in the test step in a python:3 image

 

 

 

对于.gitlab-ci.yml文件的的其他部分,请查看如下文档介绍:https://docs.gitlab.com/ce/ci/yaml/README.html

在我们当前的项目中定义了 4 个构建阶段:test、build、release、review、deploy,完整的.gitlab-ci.yml文件如下:

image:
  name: golang:1.10.3-stretch
  entrypoint: ["/bin/sh", "-c"]

# The problem is that to be able to use go get, one needs to put
# the repository in the $GOPATH. So for example if your gitlab domain
# is mydomainperso.com, and that your repository is repos/projectname, and
# the default GOPATH being /go, then you'd need to have your
# repository in /go/src/mydomainperso.com/repos/projectname
# Thus, making a symbolic link corrects this.
before_script:
  - mkdir -p "/go/src/gitlab.wangxu.com/${CI_PROJECT_NAMESPACE}"
  - ln -sf "${CI_PROJECT_DIR}" "/go/src/gitlab.wangxu.com/${CI_PROJECT_PATH}"
  - cd "/go/src/gitlab.wangxu.com/${CI_PROJECT_PATH}/"

stages:
  - test
  - build
  - release
  - review
  - deploy

test:
  stage: test
  script:
    - make test

test2:
  stage: test
  script:
    - sleep 3
    - echo "We did it! Something else runs in parallel!"

compile:
  stage: build
  script:
    # Add here all the dependencies, or use glide/govendor/...
    # to get them automatically.
    - make build
  artifacts:
    paths:
      - app

image_build:
  stage: release
  image: docker:latest
  variables:
    DOCKER_DRIVER: overlay
    DOCKER_HOST: tcp://localhost:2375
  services:
    - name: docker:17.03-dind
      command: ["--insecure-registry=harbor.wangxu.com"]
  script:
    - docker info
    - docker login -u "${CI_REGISTRY_USER}" -p "${CI_REGISTRY_PASSWORD}" "${CI_REGISTRY}"
    - docker build -t "${CI_REGISTRY_IMAGE}:latest" .
    - docker tag "${CI_REGISTRY_IMAGE}:latest" "${CI_REGISTRY}/library/${CI_REGISTRY_IMAGE}:${CI_COMMIT_REF_NAME}"
    - test ! -z "${CI_COMMIT_TAG}" && docker push "${CI_REGISTRY_IMAGE}:latest"
    - docker push "${CI_REGISTRY}/library/${CI_REGISTRY_IMAGE}:${CI_COMMIT_REF_NAME}"

deploy_review:
  image: cnych/kubectl
  stage: review
  only:
    - branches
  except:
    - tags
  environment:
    name: dev
    url: https://dev-gitlab-k8s-demo.wangxu.com
    on_stop: stop_review
  script:
    - kubectl version
    - cd manifests/
    - sed -i "s/__CI_ENVIRONMENT_SLUG__/${CI_ENVIRONMENT_SLUG}/" deployment.yaml ingress.yaml service.yaml
    - sed -i "s/__VERSION__/${CI_COMMIT_REF_NAME}/" deployment.yaml ingress.yaml service.yaml
    - |
      if kubectl apply -f deployment.yaml | grep -q unchanged; then
          echo "=> Patching deployment to force image update."
          kubectl patch -f deployment.yaml -p "{\"spec\":{\"template\":{\"metadata\":{\"annotations\":{\"ci-last-updated\":\"$(date +'%s')\"}}}}}"
      else
          echo "=> Deployment apply has changed the object, no need to force image update."
      fi
    - kubectl apply -f service.yaml || true
    - kubectl apply -f ingress.yaml
    - kubectl rollout status -f deployment.yaml
    - kubectl get all,ing -l ref=${CI_ENVIRONMENT_SLUG}

stop_review:
  image: cnych/kubectl
  stage: review
  variables:
    GIT_STRATEGY: none
  when: manual
  only:
    - branches
  except:
    - master
    - tags
  environment:
    name: dev
    action: stop
  script:
    - kubectl version
    - kubectl delete ing -l ref=${CI_ENVIRONMENT_SLUG}
    - kubectl delete all -l ref=${CI_ENVIRONMENT_SLUG}

deploy_live:
  image: cnych/kubectl
  stage: deploy
  environment:
    name: live
    url: https://live-gitlab-k8s-demo.wangxu.com
  only:
    - tags
  when: manual
  script:
    - kubectl version
    - cd manifests/
    - sed -i "s/__CI_ENVIRONMENT_SLUG__/${CI_ENVIRONMENT_SLUG}/" deployment.yaml ingress.yaml service.yaml
    - sed -i "s/__VERSION__/${CI_COMMIT_REF_NAME}/" deployment.yaml ingress.yaml service.yaml
    - kubectl apply -f deployment.yaml
    - kubectl apply -f service.yaml
    - kubectl apply -f ingress.yaml
    - kubectl rollout status -f deployment.yaml
    - kubectl get all,ing -l ref=${CI_ENVIRONMENT_SLUG}

 

上面的.gitlab-ci.yml文件中还有一些特殊的属性,如限制运行的的whenonly参数,例如only: ["tags"]表示只为创建的标签运行,更多的信息,我可以通过查看 Gitlab CI YAML 文件查看:https://docs.gitlab.com/ce/ci/yaml/README.html

由于我们在.gitlab-ci.yml文件中将应用的镜像构建完成后推送到了我们的私有仓库,而 Kubernetes 资源清单文件中使用的私有镜像,所以我们需要配置一个imagePullSecret,否则在 Kubernetes 集群中是无法拉取我们的私有镜像的:(替换下面相关信息为自己的)

# kubectl create secret docker-registry myregistry --docker-server=harbor.wangxu.com --docker-username=admin --docker-password=Harbor12345 --docker-email=314144952@qq.com -n gitlab
secret/myregistry created

 

 

在下面的 Deployment 的资源清单文件中会使用到创建的myregistry

创建k8s资源清单文件

接下来为应用创建 Kubernetes 资源清单文件,添加到代码仓库中。首先创建 Deployment 资源:(deployment.yaml)

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: gitlab-k8s-demo-__CI_ENVIRONMENT_SLUG__
  namespace: gitlab
  labels:
    app: gitlab-k8s-demo
    ref: __CI_ENVIRONMENT_SLUG__
    track: stable
spec:
  replicas: 2
  selector:
    matchLabels:
      app: gitlab-k8s-demo
      ref: __CI_ENVIRONMENT_SLUG__
  template:
    metadata:
      labels:
        app: gitlab-k8s-demo
        ref: __CI_ENVIRONMENT_SLUG__
        track: stable
    spec:
      imagePullSecrets:
        - name: myregistry
      containers:
      - name: app
        image: harbor.wangxu.com/library/gitlab-ci-k8s-demo:__VERSION__  
        #注意是harbor地址,原作者github有不一致的地方,注意看日志调整
        imagePullPolicy: Always
        ports:
        - name: http-metrics
          protocol: TCP
          containerPort: 8000
        livenessProbe:
          httpGet:
            path: /health
            port: 8000
          initialDelaySeconds: 3
          timeoutSeconds: 2
        readinessProbe:
          httpGet:
            path: /health
            port: 8000
          initialDelaySeconds: 3
          timeoutSeconds: 2

 

#注意用上面创建的 myregistry 替换 imagePullSecrets。

 

 

这是一个基本的 Deployment 资源清单的描述,像__CI_ENVIRONMENT_SLUG____VERSION__这样的占位符用于区分不同的环境,__CI_ENVIRONMENT_SLUG__将由 dev 或 live(环境名称)和__VERSION__替换为镜像标签。

为了能够连接到部署的 Pod,还需要 Service。对应的 Service 资源清单如下(service.yaml):

---
apiVersion: v1
kind: Service
metadata:
  name: gitlab-k8s-demo-__CI_ENVIRONMENT_SLUG__
  namespace: gitlab
  labels:
    app: gitlab-k8s-demo
    ref: __CI_ENVIRONMENT_SLUG__
  annotations:
    prometheus.io/scrape: "true"
    prometheus.io/port: "8000"
    prometheus.io/scheme: "http"
    prometheus.io/path: "/metrics"
spec:
  type: ClusterIP
  ports:
    - name: http-metrics
      port: 8000
      protocol: TCP
  selector:
    app: gitlab-k8s-demo
    ref: __CI_ENVIRONMENT_SLUG__

 

我们的应用程序运行8000端口上,端口名为http-metrics,如果你还记得前面我们监控的课程中应该还记得我们使用prometheus-operator为 Prometheus 创建了自动发现的配置,所以我们在annotations里面配置上上面的这几个注释后,Prometheus 就可以自动获取我们应用的监控指标数据了。

现在 Service 创建成功了,但是外部用户还不能访问到我们的应用,当然我们可以把 Service 设置成 NodePort 类型,另外一个常见的方式当然就是使用 Ingress 了,我们可以通过 Ingress 来将应用暴露给外面用户使用,对应的资源清单文件如下:(ingress.yaml)

---
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: gitlab-k8s-demo-__CI_ENVIRONMENT_SLUG__
  namespace: gitlab
  labels:
    app: gitlab-k8s-demo
    ref: __CI_ENVIRONMENT_SLUG__
  annotations:
    kubernetes.io/ingress.class: "nginx"
spec:
  rules:
  - host: __CI_ENVIRONMENT_SLUG__-gitlab-k8s-demo.wangxu.com
    http:
      paths:
      - path: /
        backend:
          serviceName: gitlab-k8s-demo-__CI_ENVIRONMENT_SLUG__
          servicePort: 8000

 

 

当然如果想配置 https 访问的话我们可以自己用 CA 证书创建一个 tls 密钥,也可以使用cert-manager来自动为我们的应用程序添加 https。

当然要通过上面的域名进行访问,还需要进行 DNS 解析的,__CI_ENVIRONMENT_SLUG__-gitlab-k8s-demo.qikqiak.com其中__CI_ENVIRONMENT_SLUG__值为 live 或 dev,所以需要创建dev-gitlab-k8s-demo.wangxu.comlive-gitlab-k8s-demo.wangxu.com 两个域名的解析。

 

我们可以使用 DNS 解析服务商的 API 来自动创建域名解析,也可以使用 Kubernetes incubator 孵化的项目 external-dns operator 来进行操作。

 

触发gitlab-CI

所需要的资源清单和.gitlab-ci.yml文件已经准备好了,我们可以小小的添加一个文件去触发下 Gitlab CI 构建:

 

touch test1
git add .
git commit -m"Testing the GitLab CI functionality #1"
git push origin master

 

 

还是失败了,我的node节点没有配置harbor的域名

  Warning  Failed     1s (x2 over 16s)   kubelet, k8s-node-2  Failed to pull image "harbor.wangxu.com/library/gitlab-ci-k8s-demo:master": rpc error: code = Unknown desc = Error response from daemon: Get https://harbor.wangxu.com/v2/: x509: certificate signed by unknown authority
  Warning  Failed     1s (x2 over 16s)   kubelet, k8s-node-2  Error: ErrImagePull
[root@k8s-master gitlab-ci-k8s-demo]#

 

[root@k8s-node-1 ~]# tail /etc/hosts

10.6.76.23 gitlab.wangxu.com harbor.wangxu.com
10.6.76.24 gitlab.wangxu.com harbor.wangxu.com

[root@k8s-node-1 ~]# vim /etc/docker/daemon.json
[root@k8s-node-1 ~]# systemctl restart docker
[root@k8s-node-1 ~]# cat /etc/docker/daemon.json
{
    "registry-mirrors": ["http://hub-mirror.c.163.com"],
     "insecure-registries": ["harbor.wangxu.com"]
}
[root@k8s-node-1 ~]#

 

 

真是被自己坑死了,再重新构建

 

 

 

 

 

整个 Pipeline 构建成功后,我们可以在项目的环境菜单下面看到多了一个环境:

 

 

 

 

 

#解析 ingress域名
10.6.76.23  dev-gitlab-k8s-demo.wangxu.com
10.6.76.24  dev-gitlab-k8s-demo.wangxu.com

 

 

 

瑕疵和没有解决的

 

 

终止操作无效啊。

 

按照设定

  - kubectl delete ing -l ref=${CI_ENVIRONMENT_SLUG}
  - kubectl delete all -l ref=${CI_ENVIRONMENT_SLUG}

 

但是无效,调试几次还是不行,不过也不影响实现了 gitlac-CI部署。有缘看到的,看是否遇到

 

posted on 2019-10-24 10:50  光阴8023  阅读(5983)  评论(0编辑  收藏  举报