kubeSourceCodeNote/apiServer/Kubernetes源码学习-APIServer-P4-APIServer的鉴权机制.md

## 前言

在上一篇[**APIServer-P3-APIServer的认证机制**](https://github.com/yinwenqin/kubeSourceCodeNote/blob/master/apiServer/Kubernetes%E6%BA%90%E7%A0%81%E5%AD%A6%E4%B9%A0-APIServer-P3-APIServer%E7%9A%84%E8%AE%A4%E8%AF%81%E6%9C%BA%E5%88%B6.md)中，讲述了请求进入后的认证过程，在通过认证之后，请求将进入鉴权环节，本篇就此展开。



## 审查请求属性

Kubernetes 仅审查以下 API 请求属性：

- **用户** - 身份验证期间提供的 `user` 字符串。
- **组** - 经过身份验证的用户所属的组名列表。
- **额外信息** - 由身份验证层提供的任意字符串键到字符串值的映射。
- **API** - 指示请求是否针对 API 资源。
- **请求路径** - 各种非资源端点的路径，如 `/api` 或 `/healthz`。
- **API 请求动词** - API 动词 `get`、`list`、`create`、`update`、`patch`、`watch`、 `proxy`、`redirect`、`delete` 和 `deletecollection` 用于资源请求。 要确定资源 API 端点的请求动词，请参阅 [确定请求动词](https://kubernetes.io/zh/docs/reference/access-authn-authz/authorization/#determine-the-request-verb)。
- **HTTP 请求动词** - HTTP 动词 `get`、`post`、`put` 和 `delete` 用于非资源请求。
- **Resource** - 正在访问的资源的 ID 或名称（仅限资源请求）- 对于使用 `get`、`update`、`patch` 和 `delete` 动词的资源请求，你必须提供资源名称。
- **子资源** - 正在访问的子资源（仅限资源请求）。
- **名字空间** - 正在访问的对象的名称空间（仅适用于名字空间资源请求）。
- **API 组** - 正在访问的 [API 组](https://kubernetes.io/zh/docs/concepts/overview/kubernetes-api/#api-groups) （仅限资源请求）。空字符串表示[核心 API 组](https://kubernetes.io/zh/docs/reference/using-api/#api-groups)



## 鉴权的描述



### 鉴权策略分类

目前支持6种鉴权策略，每种鉴权策略对应一个鉴权器，使用的鉴权策略需要在APIServer启动时以参数`--authorization-mode`的形式指定，多种策略同时指定时使用','号连接：

策略分类有：

- `--authorization-mode=ABAC` 基于属性的访问控制（ABAC）模式允许你 使用本地文件配置策略。
- `--authorization-mode=RBAC` 基于角色的访问控制（RBAC）模式允许你使用 Kubernetes API 创建和存储策略。
- `--authorization-mode=Webhook` WebHook 是一种 HTTP 回调模式，允许你使用远程 REST 端点管理鉴权。
- `--authorization-mode=Node` 节点鉴权是一种特殊用途的鉴权模式，专门对 kubelet 发出的 API 请求执行鉴权。
- `--authorization-mode=AlwaysDeny` 该标志阻止所有请求。仅将此标志用于测试。
- `--authorization-mode=AlwaysAllow` 此标志允许所有请求。仅在你不需要 API 请求 的鉴权时才使用此标志。

与上一篇的认证模块不同的是，当配置多个鉴权模块时，鉴权模块按**顺序**检查，靠前的模块具有更高的优先级来允许或拒绝请求。

来看看现有的kubeadm部署集群启用的鉴权策略:

![](https://mycloudn.upweto.top/20210203155452.png)

可以看到，默认启用了Node授权和RBAC授权模块。



### 鉴权结果

对于每一个请求的鉴权结果，有专门为其设计的健全结果描述结构体，如下:

`vendor/k8s.io/apiserver/pkg/authorization/authorizer/interfaces.go:148`

```go
type Decision int

const (
	// 拒绝
	DecisionDeny Decision = iota
	// 允许，则鉴权流程视为成功，请求顺利进入
	DecisionAllow
	// 无操作，进入下一个鉴权模块，相当于pass
	DecisionNoOpinion
)
```



### 鉴权接口方法

`vendor/k8s.io/apiserver/pkg/authorization/authorizer/interfaces.go:69`

```go
// Authorizer makes an authorization decision based on information gained by making
// zero or more calls to methods of the Attributes interface.  It returns nil when an action is
// authorized, otherwise it returns an error.
type Authorizer interface {
	Authorize(a Attributes) (authorized Decision, reason string, err error)
}

```

所有的鉴权模块(鉴权器)都要实现这个`Authorize`方法,返回鉴权结果。



### 规则解析器

规则解析器可以根据认证之后所得到的用户信息，获取该用户对应的资源对象的操作权限。

```go
// RuleResolver provides a mechanism for resolving the list of rules that apply to a given user within a namespace.
type RuleResolver interface {
	// RulesFor get the list of cluster wide rules, the list of rules in the specific namespace, incomplete status and errors.
	RulesFor(user user.Info, namespace string) ([]ResourceRuleInfo, []NonResourceRuleInfo, bool, error)
}
```

以这里的返回值类型`ResourceRuleInfo`为例，默认的`DefaultResourceRuleInfo`结构体是这样的:

`vendor/k8s.io/apiserver/pkg/authorization/authorizer/rule.go:31`

```go
// DefaultResourceRuleInfo holds information that describes a rule for the resource
type DefaultResourceRuleInfo struct {
	Verbs         []string
	APIGroups     []string
	Resources     []string
	ResourceNames []string
}
```

例如对pod资源的任意操作权限的描述可以描述为:

```go
DefaultResourceRuleInfo{
	Verbs         []string{"*"}
	APIGroups     []string{"*"}
	Resources     []string{"pod"}
}
```

这个DefaultResourceRuleInfo对象描述的规则是，允许对所有api group 的pod资源进行的所有类型的操作，包括{"get", "list", "update", "patch","create", "delete", "watch", "deletecollection"}操作。



### 鉴权流程图

![](https://mycloudn.upweto.top/20210203180748.png)



## 鉴权器

`AlwaysAllow`和`AlwaysDeny`这两种鉴权器很少使用，就不看了，直接略过.

### ABAC鉴权器

#### 简介

基于属性的访问控制（Attribute-based access control - ABAC）定义了访问控制范例，其中通过使用将属性组合在一起的策略来向用户授予访问权限。

启用ABAC鉴权器需要额外增加一个`--authorization-policy-file=SOME_FILENAME`参数，指定一个json格式的文件预设鉴权策略，是一种静态的权限配置方式。json格式样例如下：

```json
// 授予pod资源的任意操作权限给用户podManager
{
    "apiVersion":"abac.authorization.kubernetes.io/v1beta1",
    "kind":"Policy",
    "spec":{
        "user":"podManager",
        "namespace":"*",
        "resource":"pods",
        "readonly":true
    }
}

```



#### 代码实现

`pkg/auth/authorizer/abac/abac.go:224`

```go
// Authorizer implements authorizer.Authorize
func (pl policyList) Authorize(a authorizer.Attributes) (authorizer.Decision, string, error) {
	for _, p := range pl {
		if matches(*p, a) {
			return authorizer.DecisionAllow, "", nil
		}
	}
	return authorizer.DecisionNoOpinion, "No policy matched.", nil
}
```

--> `pkg/auth/authorizer/abac/abac.go:117`

```go
func matches(p abac.Policy, a authorizer.Attributes) bool {
	if subjectMatches(p, a.GetUser()) {
    // 操作类型与规则匹配
		if verbMatches(p, a) {
			// Resource and non-resource requests are mutually exclusive, at most one will match a policy
      // 资源类型与规则匹配(包含namespace/APIGroup/Resource)
			if resourceMatches(p, a) {
				return true
			}
      // 针对非资源对象的操作匹配(请求路径匹配)
			if nonResourceMatches(p, a) {
				return true
			}
		}
	}
	return false
}

```



### RBAC鉴权器

#### 简介

基于角色（Role）的访问控制（RBAC）是一种基于组织中用户的角色来调节控制对 计算机或网络资源的访问的方法。

RBAC 鉴权机制使用 `rbac.authorization.k8s.io` [API 组](https://kubernetes.io/zh/docs/concepts/overview/kubernetes-api/#api-groups) 来驱动鉴权决定，允许你通过 Kubernetes API 动态配置策略。

通过创建`Role` 或` ClusterRole`来描述具体的资源授权策略，再通过创建RoleBinding/ClusterRoleBinding将策略绑定到用户/群组/服务上。

RBAC模式的详细描述和使用样例请参考我之前的文章：

[k8s(十四)、RBAC权限控制](https://blog.upweto.top/2018/12/06/k8s(%E5%8D%81%E5%9B%9B)%E3%80%81RBAC%E6%9D%83%E9%99%90%E6%8E%A7%E5%88%B6.html)



#### 代码实现

`plugin/pkg/auth/authorizer/rbac/rbac.go:74`

```go
func (r *RBACAuthorizer) Authorize(requestAttributes authorizer.Attributes) (authorizer.Decision, string, error) {
	ruleCheckingVisitor := &authorizingVisitor{requestAttributes: requestAttributes}
  // 规则解析器解析请求的属性，返回鉴权结果，判断匹配用的是ruleCheckingVisitor.visit方法
	r.authorizationRuleResolver.VisitRulesFor(requestAttributes.GetUser(), requestAttributes.GetNamespace(), ruleCheckingVisitor.visit)
	if ruleCheckingVisitor.allowed {
		return authorizer.DecisionAllow, ruleCheckingVisitor.reason, nil
	}

	...

	reason := ""
	if len(ruleCheckingVisitor.errors) > 0 {
		reason = fmt.Sprintf("RBAC: %v", utilerrors.NewAggregate(ruleCheckingVisitor.errors))
	}
	return authorizer.DecisionNoOpinion, reason, nil
}
```

对比规则和请求属性，返回true or false的visit方法：

`plugin/pkg/auth/authorizer/rbac/rbac.go:62`

```go
func (v *authorizingVisitor) visit(source fmt.Stringer, rule *rbacv1.PolicyRule, err error) bool {
	if rule != nil && RuleAllows(v.requestAttributes, rule) {
    // 请求与规则匹配，则鉴权成功
		v.allowed = true
		v.reason = fmt.Sprintf("RBAC: allowed by %s", source.String())
    // 返回false是为了提前break鉴权过程
		return false
	}
	if err != nil {
		v.errors = append(v.errors, err)
	}
	return true
}
```



获取规则并调用visit方法的是`VisitRulesFor`接口方法，找一下`VisitRulesFor`方法:

`pkg/registry/rbac/validation/rule.go:178`

```go

func (r *DefaultRuleResolver) VisitRulesFor(user user.Info, namespace string, visitor func(source fmt.Stringer, rule *rbacv1.PolicyRule, err error) bool) {
  // 先拿到所有的ClusterRoleBinding对象，ClusterRoleBinding资源是cluster级别的，不区分命名空间
	if clusterRoleBindings, err := r.clusterRoleBindingLister.ListClusterRoleBindings(); err != nil {
		if !visitor(nil, nil, err) {
			return
		}
	} else {
		sourceDescriber := &clusterRoleBindingDescriber{}
		for _, clusterRoleBinding := range clusterRoleBindings {
      // clusterRoleBinding.Subjects指定的绑定的用户对象，对比请求的所属用户，不匹配则continue
			subjectIndex, applies := appliesTo(user, clusterRoleBinding.Subjects, "")
			if !applies {
				continue
			}
			rules, err := r.GetRoleReferenceRules(clusterRoleBinding.RoleRef, "")
			if err != nil {
				if !visitor(nil, nil, err) {
					return
				}
				continue
			}
			sourceDescriber.binding = clusterRoleBinding
			sourceDescriber.subject = &clusterRoleBinding.Subjects[subjectIndex]
			for i := range rules {
        // visit方法返回false是代表鉴权成功了，提前break鉴权过程
				if !visitor(sourceDescriber, &rules[i], nil) {
					return
				}
			}
		}
	}
	
  // 如果指定了namespace，再取命名空间级别的roleBinding资源对象，重复一次上面的过程
	if len(namespace) > 0 {
		if roleBindings, err := r.roleBindingLister.ListRoleBindings(namespace); err != nil {
			if !visitor(nil, nil, err) {
				return
			}
		} else {
			sourceDescriber := &roleBindingDescriber{}
			for _, roleBinding := range roleBindings {
				subjectIndex, applies := appliesTo(user, roleBinding.Subjects, namespace)
				if !applies {
					continue
				}
				rules, err := r.GetRoleReferenceRules(roleBinding.RoleRef, namespace)
				if err != nil {
					if !visitor(nil, nil, err) {
						return
					}
					continue
				}
				sourceDescriber.binding = roleBinding
				sourceDescriber.subject = &roleBinding.Subjects[subjectIndex]
				for i := range rules {
					if !visitor(sourceDescriber, &rules[i], nil) {
						return
					}
				}
			}
		}
	}
}

```



#### 过程总结

参考代码片中的注释，rbac鉴权过程如下:

- 1.取到所有的clusterRoleBinding/roleBindings资源对象，遍历它们对比请求用户
- 2.对比roleBindings/clusterRoleBinding指向的用户(主体)与请求用户，相同则选中，不相同continue
- 3.对比规则与请求属性，符合则提前结束鉴权



### Node鉴权器

上面有提过，node鉴权器是专为kubelet组件设计的，按照kubeadm集群的默认配置，它是排序在第一位的鉴权器，为什么把它放在后面再讲呢，因为node鉴权器本质上也是利用了rbac鉴权器，是通过为`system:node`这个内置用户授权来实现的，来看一下。

#### 默认Node规则生成

`plugin/pkg/auth/authorizer/rbac/bootstrappolicy/policy.go:97`

```go
func NodeRules() []rbacv1.PolicyRule {
	nodePolicyRules := []rbacv1.PolicyRule{
		// Needed to check API access.  These creates are non-mutating
		rbacv1helpers.NewRule("create").Groups(authenticationGroup).Resources("tokenreviews").RuleOrDie(),
		rbacv1helpers.NewRule("create").Groups(authorizationGroup).Resources("subjectaccessreviews", "localsubjectaccessreviews").RuleOrDie(),

		// Needed to build serviceLister, to populate env vars for services
		rbacv1helpers.NewRule(Read...).Groups(legacyGroup).Resources("services").RuleOrDie(),

		// Nodes can register Node API objects and report status.
		// Use the NodeRestriction admission plugin to limit a node to creating/updating its own API object.
		rbacv1helpers.NewRule("create", "get", "list", "watch").Groups(legacyGroup).Resources("nodes").RuleOrDie(),
		rbacv1helpers.NewRule("update", "patch").Groups(legacyGroup).Resources("nodes/status").RuleOrDie(),
		rbacv1helpers.NewRule("update", "patch").Groups(legacyGroup).Resources("nodes").RuleOrDie(),

		// TODO: restrict to the bound node as creator in the NodeRestrictions admission plugin
		rbacv1helpers.NewRule("create", "update", "patch").Groups(legacyGroup).Resources("events").RuleOrDie(),

		// TODO: restrict to pods scheduled on the bound node once field selectors are supported by list/watch authorization
		rbacv1helpers.NewRule(Read...).Groups(legacyGroup).Resources("pods").RuleOrDie(),

		// Needed for the node to create/delete mirror pods.
		// Use the NodeRestriction admission plugin to limit a node to creating/deleting mirror pods bound to itself.
		rbacv1helpers.NewRule("create", "delete").Groups(legacyGroup).Resources("pods").RuleOrDie(),
		// Needed for the node to report status of pods it is running.
		// Use the NodeRestriction admission plugin to limit a node to updating status of pods bound to itself.
		rbacv1helpers.NewRule("update", "patch").Groups(legacyGroup).Resources("pods/status").RuleOrDie(),
		// Needed for the node to create pod evictions.
		// Use the NodeRestriction admission plugin to limit a node to creating evictions for pods bound to itself.
		rbacv1helpers.NewRule("create").Groups(legacyGroup).Resources("pods/eviction").RuleOrDie(),

		// Needed for imagepullsecrets, rbd/ceph and secret volumes, and secrets in envs
		// Needed for configmap volume and envs
		// Use the Node authorization mode to limit a node to get secrets/configmaps referenced by pods bound to itself.
		rbacv1helpers.NewRule("get", "list", "watch").Groups(legacyGroup).Resources("secrets", "configmaps").RuleOrDie(),
		// Needed for persistent volumes
		// Use the Node authorization mode to limit a node to get pv/pvc objects referenced by pods bound to itself.
		rbacv1helpers.NewRule("get").Groups(legacyGroup).Resources("persistentvolumeclaims", "persistentvolumes").RuleOrDie(),

		// TODO: add to the Node authorizer and restrict to endpoints referenced by pods or PVs bound to the node
		// Needed for glusterfs volumes
		rbacv1helpers.NewRule("get").Groups(legacyGroup).Resources("endpoints").RuleOrDie(),
		// Used to create a certificatesigningrequest for a node-specific client certificate, and watch
		// for it to be signed. This allows the kubelet to rotate it's own certificate.
		rbacv1helpers.NewRule("create", "get", "list", "watch").Groups(certificatesGroup).Resources("certificatesigningrequests").RuleOrDie(),
	}

	if utilfeature.DefaultFeatureGate.Enabled(features.ExpandPersistentVolumes) {
		// Use the Node authorization mode to limit a node to update status of pvc objects referenced by pods bound to itself.
		// Use the NodeRestriction admission plugin to limit a node to just update the status stanza.
		pvcStatusPolicyRule := rbacv1helpers.NewRule("get", "update", "patch").Groups(legacyGroup).Resources("persistentvolumeclaims/status").RuleOrDie()
		nodePolicyRules = append(nodePolicyRules, pvcStatusPolicyRule)
	}

	if utilfeature.DefaultFeatureGate.Enabled(features.TokenRequest) {
		// Use the Node authorization to limit a node to create tokens for service accounts running on that node
		// Use the NodeRestriction admission plugin to limit a node to create tokens bound to pods on that node
		tokenRequestRule := rbacv1helpers.NewRule("create").Groups(legacyGroup).Resources("serviceaccounts/token").RuleOrDie()
		nodePolicyRules = append(nodePolicyRules, tokenRequestRule)
	}

	// CSI
	if utilfeature.DefaultFeatureGate.Enabled(features.CSIPersistentVolume) {
		volAttachRule := rbacv1helpers.NewRule("get").Groups(storageGroup).Resources("volumeattachments").RuleOrDie()
		nodePolicyRules = append(nodePolicyRules, volAttachRule)
		if utilfeature.DefaultFeatureGate.Enabled(features.CSIDriverRegistry) {
			csiDriverRule := rbacv1helpers.NewRule("get", "watch", "list").Groups("storage.k8s.io").Resources("csidrivers").RuleOrDie()
			nodePolicyRules = append(nodePolicyRules, csiDriverRule)
		}
	}
	if utilfeature.DefaultFeatureGate.Enabled(features.KubeletPluginsWatcher) &&
		utilfeature.DefaultFeatureGate.Enabled(features.CSINodeInfo) {
		csiNodeInfoRule := rbacv1helpers.NewRule("get", "create", "update", "patch", "delete").Groups("storage.k8s.io").Resources("csinodes").RuleOrDie()
		nodePolicyRules = append(nodePolicyRules, csiNodeInfoRule)
	}

	// Node leases
	if utilfeature.DefaultFeatureGate.Enabled(features.NodeLease) {
		nodePolicyRules = append(nodePolicyRules, rbacv1helpers.NewRule("get", "create", "update", "patch", "delete").Groups("coordination.k8s.io").Resources("leases").RuleOrDie())
	}

	// RuntimeClass
	if utilfeature.DefaultFeatureGate.Enabled(features.RuntimeClass) {
		nodePolicyRules = append(nodePolicyRules, rbacv1helpers.NewRule("get", "list", "watch").Groups("node.k8s.io").Resources("runtimeclasses").RuleOrDie())
	}
	return nodePolicyRules
}
```

这里初始化了kubelet工作所需要的资源的权限，如(node/pod/cm/secret/pvc等)



#### Authorize代码实现

`plugin/pkg/auth/authorizer/node/node_authorizer.go:80`

```go
func (r *NodeAuthorizer) Authorize(attrs authorizer.Attributes) (authorizer.Decision, string, error) {
  // 判断是不是node发起的请求(所属的group是不是system:node)
	nodeName, isNode := r.identifier.NodeIdentity(attrs.GetUser())
	if !isNode {
		// reject requests from non-nodes
		return authorizer.DecisionNoOpinion, "", nil
	}
	if len(nodeName) == 0 {
		// reject requests from unidentifiable nodes
		klog.V(2).Infof("NODE DENY: unknown node for user %q", attrs.GetUser().GetName())
		return authorizer.DecisionNoOpinion, fmt.Sprintf("unknown node for user %q", attrs.GetUser().GetName()), nil
	}

	// subdivide access to specific resources
	if attrs.IsResourceRequest() {
    // 根据请求属性(路径)获取资源类型，不同类型资源不同的方式处理
		requestResource := schema.GroupResource{Group: attrs.GetAPIGroup(), Resource: attrs.GetResource()}
		switch requestResource {
		case secretResource:
			return r.authorizeReadNamespacedObject(nodeName, secretVertexType, attrs)
		case configMapResource:
			return r.authorizeReadNamespacedObject(nodeName, configMapVertexType, attrs)
		case pvcResource:
			if r.features.Enabled(features.ExpandPersistentVolumes) {
				if attrs.GetSubresource() == "status" {
					return r.authorizeStatusUpdate(nodeName, pvcVertexType, attrs)
				}
			}
			return r.authorizeGet(nodeName, pvcVertexType, attrs)
		case pvResource:
			return r.authorizeGet(nodeName, pvVertexType, attrs)
		case vaResource:
			if r.features.Enabled(features.CSIPersistentVolume) {
				return r.authorizeGet(nodeName, vaVertexType, attrs)
			}
			return authorizer.DecisionNoOpinion, fmt.Sprintf("disabled by feature gate %s", features.CSIPersistentVolume), nil
		case svcAcctResource:
			if r.features.Enabled(features.TokenRequest) {
				return r.authorizeCreateToken(nodeName, serviceAccountVertexType, attrs)
			}
			return authorizer.DecisionNoOpinion, fmt.Sprintf("disabled by feature gate %s", features.TokenRequest), nil
		case leaseResource:
			if r.features.Enabled(features.NodeLease) {
				return r.authorizeLease(nodeName, attrs)
			}
			return authorizer.DecisionNoOpinion, fmt.Sprintf("disabled by feature gate %s", features.NodeLease), nil
		case csiNodeResource:
			if r.features.Enabled(features.KubeletPluginsWatcher) && r.features.Enabled(features.CSINodeInfo) {
				return r.authorizeCSINode(nodeName, attrs)
			}
			return authorizer.DecisionNoOpinion, fmt.Sprintf("disabled by feature gates %s and %s", features.KubeletPluginsWatcher, features.CSINodeInfo), nil
		}

	}

	// Access to other resources is not subdivided, so just evaluate against the statically defined node rules
	if rbac.RulesAllow(attrs, r.nodeRules...) {
		return authorizer.DecisionAllow, "", nil
	}
	return authorizer.DecisionNoOpinion, "", nil
}
```



### WebHook鉴权器

#### 简介

与上一篇中的WebHook认证器类似，WebHook鉴权器也是依赖于集群外部的鉴权服务器，将鉴权请求POST发送给外部的鉴权服务器。

`Webhook` 模式需要一个 HTTP 配置文件，通过 `--authorization-webhook-config-file=SOME_FILENAME` 的参数声明。

配置文件的格式使用 [kubeconfig](https://kubernetes.io/zh/docs/tasks/access-application-cluster/configure-access-multiple-clusters/)。在文件中，"users" 代表着 API 服务器的 webhook，而 "cluster" 代表着远程服务。

使用 HTTPS 客户端认证的配置例子：

```yaml
# Kubernetes API 版本
apiVersion: v1
# API 对象种类
kind: Config
# clusters 代表远程服务。
clusters:
  - name: name-of-remote-authz-service
    cluster:
      # 对远程服务进行身份认证的 CA。
      certificate-authority: /path/to/ca.pem
      # 远程服务的查询 URL。必须使用 'https'。
      server: https://authz.example.com/authorize

# users 代表 API 服务器的 webhook 配置
users:
  - name: name-of-api-server
    user:
      client-certificate: /path/to/cert.pem # webhook plugin 使用 cert
      client-key: /path/to/key.pem          # cert 所对应的 key

# kubeconfig 文件必须有 context。需要提供一个给 API 服务器。
current-context: webhook
contexts:
- context:
    cluster: name-of-remote-authz-service
    user: name-of-api-server
  name: webhook
```

> 摘自官方文档[Webhook 模式](https://kubernetes.io/zh/docs/reference/access-authn-authz/webhook/)

#### 代码实现

`vendor/k8s.io/apiserver/plugin/pkg/authorizer/webhook/webhook.go:152`

```go
func (w *WebhookAuthorizer) Authorize(attr authorizer.Attributes) (decision authorizer.Decision, reason string, err error) {
	r := &authorization.SubjectAccessReview{}
	if user := attr.GetUser(); user != nil {
		r.Spec = authorization.SubjectAccessReviewSpec{
			User:   user.GetName(),
			UID:    user.GetUID(),
			Groups: user.GetGroups(),
			Extra:  convertToSARExtra(user.GetExtra()),
		}
	}

	if attr.IsResourceRequest() {
		r.Spec.ResourceAttributes = &authorization.ResourceAttributes{
			Namespace:   attr.GetNamespace(),
			Verb:        attr.GetVerb(),
			Group:       attr.GetAPIGroup(),
			Version:     attr.GetAPIVersion(),
			Resource:    attr.GetResource(),
			Subresource: attr.GetSubresource(),
			Name:        attr.GetName(),
		}
	} else {
		r.Spec.NonResourceAttributes = &authorization.NonResourceAttributes{
			Path: attr.GetPath(),
			Verb: attr.GetVerb(),
		}
	}
  // 将请求的主体/资源/操作等字段放在一个json里
	key, err := json.Marshal(r.Spec)
	if err != nil {
		return w.decisionOnError, "", err
	}
  // 从本地的缓存里取，有则不发起远端post请求了
	if entry, ok := w.responseCache.Get(string(key)); ok {
		r.Status = entry.(authorization.SubjectAccessReviewStatus)
	} else {
		var (
			result *authorization.SubjectAccessReview
			err    error
		)
		webhook.WithExponentialBackoff(w.initialBackoff, func() error {
      // 缓存里没有，则发起post请求给远端鉴权服务器
			result, err = w.subjectAccessReview.Create(r)
			return err
		})
		if err != nil {
			// An error here indicates bad configuration or an outage. Log for debugging.
			klog.Errorf("Failed to make webhook authorizer request: %v", err)
			return w.decisionOnError, "", err
		}
		r.Status = result.Status
    // 长度不超过10000则缓存结果
		if shouldCache(attr) {
			if r.Status.Allowed {
				w.responseCache.Add(string(key), r.Status, w.authorizedTTL)
			} else {
				w.responseCache.Add(string(key), r.Status, w.unauthorizedTTL)
			}
		}
	}
	switch {
    // 根据远端鉴权服务器的响应状态，返回鉴权结果
	case r.Status.Denied && r.Status.Allowed:
		return authorizer.DecisionDeny, r.Status.Reason, fmt.Errorf("webhook subject access review returned both allow and deny response")
	case r.Status.Denied:
		return authorizer.DecisionDeny, r.Status.Reason, nil
	case r.Status.Allowed:
		return authorizer.DecisionAllow, r.Status.Reason, nil
	default:
		return authorizer.DecisionNoOpinion, r.Status.Reason, nil
	}

}
```



## 总结

鉴权的流程与认证的流程大体类似，但也有所不同，例如认证器链的执行顺序是无序的，而鉴权器链的执行顺序是有序的(按参数指定的顺序)。另外鉴权器的数量没有认证器那么多，因此相对容易理解一些。