简单开发k8s多集群后台

详细的说明如何利用kubernetes api创建一个支持多集群的k8s后台，这里主要实现调用kubernetes api获取查看service服务，pod服务。下一篇再说明如果操作一个service，以及注入容器console

相关的组件
- Django (2.0.7)
- kubernetes (6.0.0)
- Django Rest Framework (3.8.2)
- Bootstrap 3.3.7
- Jquery 2.1.1
- pace 页面异步加载插件

简单开发k8s多集群后台

新建一个基于django的k8s项目,命名为k8s_web_console

简单开发k8s多集群后台

增加一个cluster应用

python manage.py startapp cluster

添加 Django Rest Framework
- 使用rest框架是方便做前后端分离，如果你是个全栈工程师，建议做前后端分离，这样方便维护。但这里为了入门快速方便，如果还要加一个node.js开发，比较麻烦，所以暂时保留这种可能，但还是用django的template方式开发。

pip install djangorestframework

在setting.py中加入rest框架

INSTALLED_APPS = [
    'django.contrib.admin',
    'django.contrib.auth',
    'django.contrib.contenttypes',
    'django.contrib.sessions',
    'django.contrib.messages',
    'django.contrib.staticfiles',
    'rest_framework',
        'cluster',
]

创建数据库以及数据表

python manage.py makemigrations
python manage.py migrate

打开cluster应用下的models.py添加一个k8s集群信息的model
```
class Cluster(models.Model):
name = models.TextField()
api = models.TextField()
token = models.TextField()

class Meta :
    db_table = "cluster"

def __str__(self):
    return self.name
```
这里是k8s的连接通过api+token的方式进行连接，当然也可以用证书认证的方式
- name : 集群的名称
- api : k8s集群的api 链接
- token ： k8s集群拥有admin权限的token

cluster应用添加一个serializers.py文件

from rest_framework import serializers
from cluster.models import Cluster

class ClusterSerializer( serializers.ModelSerializer ):
    class Meta:
        model = Cluster
        fields = '__all__'

rest 框架根据这里的配置对Cluster对象自动做序列化

cluster 的view文件添加cluster的viewset

# /api/cluster/
class ClusterViewSet( viewsets.ModelViewSet ):
    queryset = Cluster.objects.all()
    serializer_class = ClusterSerializer
    # permission_classes = (IsAuthenticated,)
    parser_classes = (JSONParser,)

rest 框架的parser有多种，这里用的是jsonparser，需要再settings.py 中添加如下配置

    REST_FRAMEWORK = {
    'DEFAULT_PARSER_CLASSES': (
        'rest_framework.parsers.JSONParser',
    )
}

编辑admin.py,将cluster注册到Django Admin后台

from django.contrib import admin
from .models import Cluster

admin.site.register(Cluster)

urls.py 中设置下路由

from cluster import views as cluster_views
router = DefaultRouter()
router.register(r'cluster', cluster_views.ClusterViewSet)

urlpatterns = [
    path( 'admin/', admin.site.urls ), 
    path( 'api/', include( router.urls ) ),
    path('api-auth/', include('rest_framework.urls', namespace='rest_framework')),
]

测试下cluster的crud，以及rest api方式
- 运行django后，打开 http://127.0.0.1:8000/admin/cluster/cluster/
- 添加集群的信息

简单开发k8s多集群后台

如果用rest 框架来看,就打开 http://127.0.0.1:8000/api/cluster/

简单开发k8s多集群后台

加入kubernetes api,开始调试kubernetes
```
pip install kubernetes
```
- cluster应用的view修改如下

from pprint import pprint
from django.shortcuts import get_object_or_404
from django.shortcuts import render
from kubernetes.client import Configuration, ApiClient, CoreV1Api
from rest_framework import viewsets
from rest_framework.decorators import action
from rest_framework.parsers import JSONParser

from cluster.models import Cluster
from .serializers import ClusterSerializer

def get_config(cluster):
    '''
    :param cluster:  k8s集群的配置对象
    :return:   返回一个config对象
    '''
    configuration = Configuration()
    configuration.verify_ssl = False
    configuration.host = cluster.api
    configuration.api_key['authorization'] = cluster.token
    configuration.api_key_prefix['authorization'] = 'Bearer'
    return configuration

def get_clients(cluster):
    '''
    :param cluster: k8s集群对象
    :return:  一个apiclient对象，一个coreV1Api对象
    '''
    api_client = ApiClient( get_config( cluster ) )
    return api_client, CoreV1Api( api_client )

# /api/cluster/
class ClusterViewSet( viewsets.ModelViewSet ):
    queryset = Cluster.objects.all()
    serializer_class = ClusterSerializer
    # permission_classes = (IsAuthenticated,)
    parser_classes = (JSONParser,)

    # /api/cluster/[pk]/list_service/
    @action( detail=True, methods=['get'] )
    def list_service(self, request, pk=None):
        cluster = get_object_or_404( Cluster, pk=pk )
        api_client, core_v1_api = get_clients( cluster )
        v1_service_list = core_v1_api.list_service_for_all_namespaces()
        # pprint(v1_service_list)
        result = {}
        items = []
        headers = ['NAMESPACE', 'NAME', 'TYPE', 'CLUSTER-IP', 'SELF LINK']

        for item in v1_service_list.items:
            items.append( [item.metadata.namespace,
                           item.metadata.name,
                           item.spec.cluster_ip, item.spec.type, item.metadata.self_link
                           ] )

        result['items'] = items
        result['headers'] = headers
        result['caption'] = "Services"
        return render( request, "table.html", {'result': result} )

    # /api/cluster/[pk]/list_pod/
    @action( detail=True, methods=['get'] )
    def list_pod(self, request, pk=None):
        cluster = get_object_or_404( Cluster, pk=pk )
        api_client, core_v1_api = get_clients( cluster )
        v1_service_list = core_v1_api.list_pod_for_all_namespaces()
        # pprint(v1_service_list)
        result = {}
        items = []
        headers = ['NAMESPACE', 'NAME', 'IP', 'NODE']

        for item in v1_service_list.items:
            items.append( [
                item.metadata.namespace,
                item.metadata.name,
                item.status.host_ip,
                item.status.pod_ip,
            ] )

        result['items'] = items
        result['headers'] = headers
        result['caption'] = "Pods"
        return render( request, "table.html", {'result': result} )

因为这里仅作了一个cluster测试，所以这里将对k8s的连接也放在这里，当如果正式的项目，应该放在一个独立的文件，创建成连接池管理，再通过引用提供到这里使用

再次声明，如果是这里为了简单，仍然用了django的template模式，当如果前后端分离开发，返回应该是序列化后的json

上面的list用到一个table.html模板，内容如下

<table class="table table-hover " id="service-table">
    <caption>{{ result.caption }}</caption>
    <thead>
    <tr>
        {% for item in result.headers %}
            <th>{{ item }}</th>
        {% endfor %}
    </tr>
    </thead>
    <tbody>
    {% for item in result.items %}
        <tr>
            {% for eny in item %}
                <td>{{ eny }}</td>
            {% endfor %}
        </tr>
    {% endfor %}
    </tbody>
    <tfoot>
    <tr>
        {% for item in result.headers %}
            <th>{{ item }}</th>
        {% endfor %}
    </tr>
    </tfoot>
</table>

项目中添加一个dashboard目录，目录下添加init.py，views.py文件
- 为了做当前页面刷新，方便管理，所以多了一个dashboard来管理，views.py的内容如下

from django.shortcuts import render
from cluster.models import Cluster

def cluster_main(request):
    result = {}
    result['clusters'] = Cluster.objects.all()
    return render(request,'main.html',{'result':result})

在urls.py 中添加一条路由,最终内容如下


from django.contrib import admin
from django.urls import path
from django.conf.urls import include
from rest_framework.routers import DefaultRouter
from cluster import views as cluster_views
from dashboard import views as dashboard_views

router = DefaultRouter()
router.register(r'cluster', cluster_views.ClusterViewSet)

urlpatterns = [
    path( 'admin/', admin.site.urls ),
    path('dash/cluster',dashboard_views.cluster_main),
    path( 'api/', include( router.urls ) ),
    path('api-auth/', include('rest_framework.urls', namespace='rest_framework')),
]

dashboard 中用了一个main.html模板，使用的是bootstrap+jquery+pace，实现当前页面刷新:

{% load static %}
<!DOCTYPE html>
<html>
<head>
    <meta charset="utf-8">
    <title>K8s 多集群管理测试</title>
    <link rel="stylesheet" href="https://cdn.bootcss.com/bootstrap/3.3.7/css/bootstrap.min.css">
    <link rel="stylesheet"
          href="https://cdn.jsdelivr.net/npm/[email protected]/src/messagebox.css">
    <link rel="stylesheet" href="{% static 'plugins/pace/pace.min.css' %}">
    <script src="https://cdn.bootcss.com/jquery/2.1.1/jquery.min.js"></script>
    <script src="https://cdn.bootcss.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>
    <script src="http://mindmup.github.io/editable-table/mindmup-editabletable.js"></script>
    <script src="https://cdn.jsdelivr.net/npm/[email protected]/src/messagebox.min.js"></script>
    <script src="{% static 'plugins/pace/pace.min.js' %}"></script>
</head>
<body>
<div class="container">
    <nav class="navbar navbar-default" role="navigation">
        <div class="container-fluid">
            <div class="navbar-header">
                <a class="navbar-brand" href="#">K8S Dashboard</a>
            </div>
            <form class="navbar-form navbar-left pull-right" role="search">
                <div class="form-group">
                    <label for="k8s-api"></label>
                    <select class=" form-control" id="k8s-api">
                        <option value="#">选择集群</option>
                        {% for cluster in result.clusters %}
                            <option value="{{ cluster.id }}">{{ cluster.name }}</option>
                        {% endfor %}
                    </select>
                </div>
            </form>
            <ul class="nav navbar-nav navbar-right">
                <li class="dropdown">
                    <a href="#" class="dropdown-toggle" data-toggle="dropdown">
                        选择类型 <b class="caret"></b>
                    </a>
                    <ul class="dropdown-menu">
                        <li><a href="javascript:void(0);" onclick="load_table('/list_pod/')">Pod</a></li>
                        <li class="divider"></li>
                        <li><a href="javascript:void(0);" onclick="load_table('/list_service/')">Service</a></li>
                    </ul>
                </li>
            </ul>
        </div>
    </nav>
    <div id="table-div"></div>
</div>

<script type="application/javascript">
    $(document).ajaxStart(function () {
        Pace.restart()
    });

    function load_table(url_method) {

        console.log(url_method);

        k8s_select = $("#k8s-api");

        k8s_id = k8s_select.val();
        if (k8s_id === "#") {
            $.MessageBox({
                buttonDone: "OK",
                message: "<center>先选择一个集群</center>",
                queue: true,
                speed: 200  ,
                top: "25%"   ,
            });
            return null;
        }
        url = '/api/cluster/' + k8s_id + url_method;
        $.ajax({
            url: url,
            success: function (ex) {
                console.log("success");
                $('#table-div').html(ex);
            },
            error: function (ex) {
                console.log(ex);
            }
        })

    }
</script>
</body>
</html>

好了，完成这些后，终于做完了，现在可以测试下，完成的页面就是上面的显示图片，下面来讲讲kubernetes api的使用
- k8s api 的主页 https://github.com/kubernetes-client/python
- 主要的操作接口： https://github.com/kubernetes-client/python/blob/master/kubernetes/docs/CoreV1Api.md
如果想要连接k8s接口并操作，需要的步骤
- 创建一个Configuration，这是k8s的连接配置信息，这里提供api接口跟token进行连接
```
configuration = Configuration()
configuration.verify_ssl = False
configuration.host = cluster.api
configuration.api_key['authorization'] = cluster.token
configuration.api_key_prefix['authorization'] = 'Bearer'
```
- 根据config创建一个CoreV1Api，这个类下包含多个操作k8s的接口方法
  api_client = ApiClient( configuration )
- 如果想要查看所有的service，利用接口list_service_for_all_namespaces

简单开发k8s多集群后台

本文的源码下载目录:http://down.51cto.com/data/2450296

简单开发k8s多集群后台

猜你喜欢