Step-by-step instructions for performing operations with Kubernetes.

Documentation for Kubernetes v1.8 is no longer actively maintained. The version you are currently viewing is a static snapshot. For up-to-date documentation, see the latest version.

Edit This Page


Kubernetes Audit provides a security-relevant chronological set of records documenting the sequence of activities that have affected system by individual users, administrators or other components of the system. It allows cluster administrator to answer the following questions:

Legacy Audit

Kubernetes audit is part of Kube-apiserver logging all requests processed by the server. Each audit log entry contains two lines:

  1. The request line containing a unique ID to match the response and request metadata, such as the source IP, requesting user, impersonation information, resource being requested, etc.
  2. The response line containing a unique ID matching the request line and the response code.

Example output for admin user listing pods in the default namespace:

2017-03-21T03:57:09.106841886-04:00 AUDIT: id="c939d2a7-1c37-4ef1-b2f7-4ba9b1e43b53" ip="" method="GET" user="admin" groups="\"system:masters\",\"system:authenticated\"" as="<self>" asgroups="<lookup>" namespace="default" uri="/api/v1/namespaces/default/pods"
2017-03-21T03:57:09.108403639-04:00 AUDIT: id="c939d2a7-1c37-4ef1-b2f7-4ba9b1e43b53" response="200"

Note that Kubernetes 1.8 has switched to use the advanced structured audit log by default. To fallback to this legacy audit, disable the advanced auditing feature using the AdvancedAuditing feature gate on the kube-apiserver:



Kube-apiserver provides the following options which are responsible for configuring where and how audit logs are handled:

If an audit log file already exists, Kubernetes appends new audit logs to that file. Otherwise, Kubernetes creates an audit log file at the location you specified in audit-log-path. If the audit log file exceeds the size you specify in audit-log-maxsize, Kubernetes will rename the current log file by appending the current timestamp on the file name (before the file extension) and create a new audit log file. Kubernetes may delete old log files when creating a new log file; you can configure how many files are retained and how old they can be by specifying the audit-log-maxbackup and audit-log-maxage options.

Advanced audit

Kubernetes 1.7 expands auditing with experimental functionality such as event filtering and a webhook for integration with external systems. Kubernetes 1.8 upgrades the advanced audit feature to beta, and some backward incompatible changes have been committed.

AdvancedAuditing is customizable in two ways. Policy, which determines what’s recorded, and backends, which persist records. Backend implementations include logs files and webhooks.

The structure of audit events changes when enabling the AdvancedAuditing feature flag. This includes some cleanups, such as the method reflecting the verb evaluated by the authorization layer instead of the HTTP verb. Also, instead of always generating two events per request, events are recorded with an associated “stage”. The known stages are:

Audit Policy

Audit policy is a document defining rules about what events should be recorded. The policy is passed to the kube-apiserver using the --audit-policy-file flag.


If AdvancedAuditing is enabled and this flag is omitted, no events are logged.

The policy file holds rules that determine the level of an event. Known audit levels are:

When an event is processed, it’s compared against the list of rules in order. The first matching rule sets the audit level of the event. The audit policy is defined by the audit.k8s.io API group. Some new fields are supported in beta version, like resourceNames and omitStages.

In Kubernetes 1.8 kind and apiVersion along with rules must be provided in the audit policy file. A policy file with 0 rules, or a policy file that doesn’t provide a valid apiVersion and kind value will be treated as illegal.

Some example audit policy files:

apiVersion: audit.k8s.io/v1beta1  #this is required in Kubernetes 1.8
kind: Policy
  # Don't log watch requests by the "system:kube-proxy" on endpoints or services
  - level: None
    users: ["system:kube-proxy"]
    verbs: ["watch"]
    - group: "" # core API group
      resources: ["endpoints", "services"]

  # Don't log authenticated requests to certain non-resource URL paths.
  - level: None
    userGroups: ["system:authenticated"]
    - "/api*" # Wildcard matching.
    - "/version"

  # Log the request body of configmap changes in kube-system.
  - level: Request
    - group: "" # core API group
      resources: ["configmaps"]
    # This rule only applies to resources in the "kube-system" namespace.
    # The empty string "" can be used to select non-namespaced resources.
    namespaces: ["kube-system"]

  # Log configmap and secret changes in all other namespaces at the Metadata level.
  - level: Metadata
    - group: "" # core API group
      resources: ["secrets", "configmaps"]

  # Log all other resources in core and extensions at the Request level.
  - level: Request
    - group: "" # core API group
    - group: "extensions" # Version of group should NOT be included.

  # A catch-all rule to log all other requests at the Metadata level.
  - level: Metadata

The next audit policy file shows new features introduced in Kubernetes 1.8:

apiVersion: audit.k8s.io/v1beta1
kind: Policy
  # Log pod changes at Request level
  - level: Request
    - group: ""
      # Resource "pods" no longer matches requests to any subresource of pods,
      # This behavior is consistent with the RBAC policy.
      resources: ["pods"]
  # Log "pods/log", "pods/status" at Metadata level
  - level: Metadata
    - group: ""
      resources: ["pods/log", "pods/status"]

  # Don't log requests to a configmap called "controller-leader"
  - level: None
    - group: ""
      resources: ["configmaps"]
      resourceNames: ["controller-leader"]

  # A catch-all rule to log all other requests at the Metadata level.
  # For this rule we use "omitStages" to omit events at "ReqeustReceived" stage.
  # Events in this stage will not be sent to backend.
  - level: Metadata
      - "RequestReceived"

You can use a minimal audit policy file to log all requests at the Metadata level:

# Log all requests at the Metadata level.
apiVersion: audit.k8s.io/v1beta1
kind: Policy
- level: Metadata

The audit profile used by GCE should be used as reference by admins constructing their own audit profiles.

Audit backends

Audit backends implement strategies for emitting events. The kube-apiserver provides a logging and webhook backend.

Each request to the API server can generate multiple events, one when the request is received, another when the response is sent, and additional events for long running requests (such as watches). The ID of events will be the same if they were generated from the same request.

The event format is defined by the audit.k8s.io API group. The v1alpha1 format of this API can be found here with more details about the exact fields captured.

Log backend

The behavior of the --audit-log-path flag changes when enabling the AdvancedAuditing feature flag. All generated events defined by --audit-policy-file are recorded in structured json format:


In alpha version, objectRef.apiVersion holds both the api group and version. In beta version these were break out into objectRef.apiGroup and objectRef.apiVersion.

Starting from Kubernetes 1.8, structured json format is used for log backend by default. Use the following option to switch log to legacy format:


With legacy format, events are formatted as follows:

2017-09-05T06:08:19.885328047-04:00 AUDIT: id="c28a95ad-f9dd-47e1-a617-b6dc152db95f" stage="RequestReceived" ip="" method="list" user="kubecfg" groups="\"system:masters\",\"system:authenticated\"" as="<self>" asgroups="<lookup>" namespace="default" uri="/apis/rbac.authorization.k8s.io/v1/namespaces/default/roles" response="<deferred>"
2017-09-05T06:08:19.885328047-04:00 AUDIT: id="c28a95ad-f9dd-47e1-a617-b6dc152db95f" stage="ResponseComplete" ip="" method="list" user="kubecfg" groups="\"system:masters\",\"system:authenticated\"" as="<self>" asgroups="<lookup>" namespace="default" uri="/apis/rbac.authorization.k8s.io/v1/namespaces/default/roles" response="200"

Logged events omit the request and response bodies. The Request and RequestResponse levels are equivalent to Metadata for legacy format. This legacy format of advanced audit is different from the Legacy Audit discussed above, such as changes to the method values and the introduction of a “stage” for each event.

Webhook backend

The audit webhook backend can be used to have kube-apiserver send audit events to a remote service. The webhook requires the AdvancedAuditing feature flag and is configured using the following command line flags:


audit-webhook-mode controls buffering strategies used by the webhook. Known modes are:

The webhook config file uses the kubeconfig format to specify the remote address of the service and credentials used to connect to it.

# clusters refers to the remote service.
  - name: name-of-remote-audit-service
      certificate-authority: /path/to/ca.pem  # CA for verifying the remote service.
      server: https://audit.example.com/audit # URL of remote service to query. Must use 'https'.

# users refers to the API server's webhook configuration.
  - name: name-of-api-server
      client-certificate: /path/to/cert.pem # cert for the webhook plugin to use
      client-key: /path/to/key.pem          # key matching the cert

# kubeconfig files require a context. Provide one for the API server.
current-context: webhook
- context:
    cluster: name-of-remote-audit-service
    user: name-of-api-sever
  name: webhook

Events are POSTed as a JSON serialized EventList. An example payload:

    "apiVersion": "audit.k8s.io/v1beta1",
    "items": [
            "auditID": "24f30caf-d7d4-45d5-b7bd-e7af300d7886",
            "level": "Metadata",
            "metadata": {
                "creationTimestamp": null
            "objectRef": {
                "apiGroup": "rbac.authorization.k8s.io",
                "apiVersion": "v1",
                "name": "jane",
                "namespace": "default",
                "resource": "roles"
            "requestURI": "/apis/rbac.authorization.k8s.io/v1/namespaces/default/roles/jane",
            "responseStatus": {
                "code": 200,
                "metadata": {}
            "sourceIPs": [
            "stage": "ResponseComplete",
            "timestamp": "2017-09-05T10:20:24Z",
            "user": {
                "groups": [
                "username": "kubecfg"
            "verb": "get"
    "kind": "EventList",
    "metadata": {}


Audit-Id is a unique ID for each http request to kube-apiserver. The ID of events will be the same if they were generated from the same request. Starting from Kubernetes 1.8, if an audit event is generated for the request, kube-apiserver will respond with an Audit-Id in the HTTP header. Note that for some special requests like kubectl exec, kubectl attach, kube-apiserver works like a proxy, no Audit-Id will be returned even if audit events are recorded.

Log Collector Examples

Use fluentd to collect and distribute audit events from log file

Fluentd is an open source data collector for unified logging layer. In this example, we will use fluentd to split audit events by different namespaces. Note that this example requries json format output support in Kubernetes 1.8.

  1. install fluentd, fluent-plugin-forest and fluent-plugin-rewrite-tag-filter in the kube-apiserver node
  2. create a config file for fluentd

    $ cat <<EOF > /etc/fluentd/config
    # fluentd conf runs in the same host with kube-apiserver
        @type tail
        # audit log path of kube-apiserver
        path /var/log/audit
        pos_file /var/log/audit.pos
        format json
        time_key time
        time_format %Y-%m-%dT%H:%M:%S.%N%z
        tag audit
    <filter audit>
        type record_transformer
         namespace ${record["objectRef"].nil? ? "none":(record["objectRef"]["namespace"].nil? ?  "none":record["objectRef"]["namespace"])}
    <match audit>
        # route audit according to namespace element in context
        @type rewrite_tag_filter
        rewriterule1 namespace ^(.+) ${tag}.$1
    <filter audit.**>
       @type record_transformer
       remove_keys namespace
    <match audit.**>
        @type forest
        subtype file
        remove_prefix audit
            time_slice_format %Y%m%d%H
            compress gz
            path /var/log/audit-${tag}.*.log
            format json
            include_time_key true
  3. start fluentd

    $ fluentd -c /etc/fluentd/config  -vv
  4. start kube-apiserver with the following options:

    --audit-policy-file=/etc/kubernetes/audit-policy.yaml --audit-log-path=/var/log/kube-audit --audit-log-format=json
  5. check audits for different namespaces in /var/log/audit-*.log

Use logstash to collect and distribute audit events from webhook backend

Logstash is an open source, server-side data processing tool. In this example, we will use logstash to collect audit events from webhook backend, and save events of different users into different files.

  1. install logstash
  2. create config file for logstash

    $ cat <<EOF > /etc/logstash/config
            #TODO, figure out a way to use kubeconfig file to authenticate to logstash
            # Webhook audit backend sends several events together with EventList
            # split each event here.
            # We only need event subelement, remove others.
            remove_field=>[headers, metadata, apiVersion, "@timestamp", kind, "@version", host]
            rename => {items=>event}
            # Audit events from different users will be saved into different files.
  3. start logstash

    $ bin/logstash -f /etc/logstash/config --path.settings /etc/logstash/
  4. create a kubeconfig file for kube-apiserver webhook audit backend

    $ cat <<EOF > /etc/kubernetes/audit-webhook-kubeconfig
    apiVersion: v1
    - cluster:
        server: http://<ip_of_logstash>:8888
      name: logstash
    - context:
        cluster: logstash
        user: ""
      name: default-context
    current-context: default-context
    kind: Config
    preferences: {}
    users: []
  5. start kube-apiserver with the following options:

    --audit-policy-file=/etc/kubernetes/audit-policy.yaml --audit-webhook-config-file=/etc/kubernetes/audit-webhook-kubeconfig
  6. check audits in logstash node’s directories /var/log/kube-audit-*/audit

Note that in addition to file output plugin, logstash has a variety of outputs that let users route data where they want. For example, users can emit audit events to elasticsearch plugin which supports full-text search and analytics.


Create an Issue Edit this Page