• kyverno VS gateKeeper


    kyverno VS gateKeeper

    概述

    这两组开源工具都是是基于kubernetes 的webhook机制,支持validatingwebhook和mutatingwebhook。整体思路上是一样的,都是针对资源的字段,如标签、镜像等来设置规则,在对kubernetes资源的控制范围和粒度上,二者可以看作是一样的。

    kyverno

    kyverno 的架构如下,它是基于kubernetes 资源的一种策略执行器,主要基于kubernetes资源的标签和spec字段制定规则,规则支持简单的条件判断,逻辑与、或、非。支持如下功能:

    • 支持集群级别和命名空间级别的策略
    • 支持审计日志功能
    • 有一个官方的UI
    • 支持kubernetes原生资源和CRD
    • 支持如下规则类型:
      validate:规则校验,最常用的类型
      mutate:支持修改现有资源
      generate:支持生成新的资源
      verifyImages:校验镜像签名

    例子

    如下策略表示拒绝没有cluster-admin clusterRoles的用户删除带app.kubernetes.io/managed-by: kyverno 标签的对象

    apiVersion: kyverno.io/v1
    kind: ClusterPolicy
    metadata:
      name: deny-deletes
    spec:
      validationFailureAction: enforce
      background: false
      rules:
      - name: block-deletes-for-kyverno-resources
        match:
          resources:
            selector:
              matchLabels:
                app.kubernetes.io/managed-by: kyverno
        exclude:
          clusterRoles:
          - cluster-admin
        validate:
          message: "Deleting {{request.oldObject.kind}}/{{request.oldObject.metadata.name}} is not allowed"
          deny:
            conditions:
            - key: "{{request.operation}}"
              operator: In
              value:
              - DELETE
    

    由于kyverno 建立在kubernetes之上,其策略决策和策略执行也是基于kubernetes的资源,因此也限制了其使用场景,如对接image registries, Active Directory/LDAP directories等第三方验证服务,而gatekeeper就可以支持就这些场景。

    此外由于它使用类yaml的方式来表达策略的,因此其使用起来比较笨拙。

    优点就是使用的配置比较简单,相比于gateKeeper来说入手比较简单,维护成本低。

    gateKeeper

    例子

    gateKeeper的规则配置要分为两步,首先创建ConstraintTemplate,再创建constraint

    首先需要创建一个模板ConstraintTemplate,下面模板用于要求所有资源中必须存在constraint 所要求的标签

    apiVersion: templates.gatekeeper.sh/v1beta1
    kind: ConstraintTemplate
    metadata:
      name: k8srequiredlabels
    spec:
      crd:
        spec:
          names:
            kind: K8sRequiredLabels
          validation:
            # Schema for the `parameters` field
            openAPIV3Schema:
              properties:
                labels:
                  type: array
                  items: string
      targets:
        - target: admission.k8s.gatekeeper.sh
          rego: |
            package k8srequiredlabels
    
            violation[{"msg": msg, "details": {"missing_labels": missing}}] {
              provided := {label | input.review.object.metadata.labels[label]}
              required := {label | label := input.parameters.labels[_]}
              missing := required - provided
              count(missing) > 0
              msg := sprintf("you must provide labels: %v", [missing])
            }
    

    然后创建一个constraints,并指定上面的K8sRequiredLabels模板,要求所有命名空间资源中必须有gatekeeper标签

    apiVersion: constraints.gatekeeper.sh/v1beta1
    kind: K8sRequiredLabels
    metadata:
      name: ns-must-have-gk
    spec:
      match:
        kinds:
          - apiGroups: [""]
            kinds: ["Namespace"]
      parameters:
        labels: ["gatekeeper"]
    

    对比表

    Features/Capabilities Gatekeeper Kyverno
    Validation
    Mutation ✓* beta
    Generation X
    Policy as native resources
    Metrics exposed
    OpenAPI validation schema (kubectl explain) X
    High Availability ✓*
    API object lookup
    CLI with test ability ✓**
    Policy audit ability
    Self-service reports X

    * Alpha status

    ** Separate CLI

    Community/Ecosystem Gatekeeper Kyverno
    CNCF status Graduated (OPA) Sandbox
    Partner ecosystem adoption*
    GitHub status (stars, forks, releases, commits) 1,832, 349, 46, 630 1,063, 122, 82, 3,326
    Community traction**
    Policy sample library

    * Not well defined. 相比Kyverno来说,Gatekeeper 的采纳意向更多,但具体不详.

    ** No objective measurement exists. 考虑到社区的存在时间,Gatekeeper 可能更具吸引力.

    Meta/Misc Gatekeeper Kyverno
    Programming required X
    Use outside Kubernetes X
    Birth (Age as of June 2021) July 2017 (3 years, 11 months) May 2019 (2 years, 1 month)
    Origin company Styra (OPA) Nirmata
    Documentation maturity ◗*

    * Not totally objective with direct comparison being difficult. Assessment made based on Gatekeeper project/functionality and not maturity level of Rego enablement materials/literature.

    参考

  • 相关阅读:
    springcloud--riboon快速搭建
    数据分析第十三讲:数据可视化入门(二)
    Elasticsearch:二进制数据类型 - binary field
    单元测试和集成测试的区别
    Vue3 源码解读系列(七)——侦听器
    Java基于 SpringBoot+Vue 的游戏分享网站
    layer弹出框详细说明
    YOLOv7模型训练
    流畅的Python读书笔记(一)序列:列表推导式和生成器表达式
    软件测试流程分享
  • 原文地址:https://www.cnblogs.com/charlieroro/p/15829201.html