Kyverno is a policy engine designed for Kubernetes.
Kyverno supports declarative validation, mutation, and generation of resource configurations using policies written as Kubernetes resources.
Kyverno can be used to scan existing workloads for best practices, or can be used to enforce best practices by blocking or mutating API requests.Kyverno allows cluster adminstrators to manage environment specific configurations independently of workload configurations and enforce configuration best practices for their clusters.
Kyverno policies are Kubernetes resources that can be written in YAML or JSON. Kyverno policies can validate, mutate, and generate any Kubernetes resources.
Kyverno runs as a dynamic admission controller in a Kubernetes cluster. Kyverno receives validating and mutating admission webhook HTTP callbacks from the kube-apiserver and applies matching policies to return results that enforce admission policies or reject requests.
Kyverno policies can match resources using the resource kind, name, and label selectors. Wildcards are supported in names.
Mutating policies can be written as overlays (similar to Kustomize) or as a JSON Patch. Validating policies also use an overlay style syntax, with support for pattern matching and conditional (if-then-else) processing.
Policy enforcement is captured using Kubernetes events. Kyverno also reports policy violations for existing resources.
NOTE : Your Kubernetes cluster version must be above v1.14 which adds webook timeouts. To check the version, enter kubectl version
.
Install Kyverno:
kubectl create -f https://github.com/nirmata/kyverno/raw/master/definitions/install.yaml
You can also install using the Helm chart. As a next step, import sample policies and learn about writing policies. You can test policies using the Kyverno cli. See docs for more details.
This policy requires that all pods have CPU and memory resource requests and limits:
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: check-cpu-memory
spec:
# `enforce` blocks the request. `audit` reports violations
validationFailureAction: enforce
rules:
- name: check-pod-resources
match:
resources:
kinds:
- Pod
validate:
message: "CPU and memory resource requests and limits are required"
pattern:
spec:
containers:
# 'name: *' selects all containers in the pod
- name: "*"
resources:
limits:
# '?' requires 1 alphanumeric character and '*' means that
# there can be 0 or more characters. Using them together
# e.g. '?*' requires at least one character.
memory: "?*"
cpu: "?*"
requests:
memory: "?*"
cpu: "?*"
This policy prevents users from changing default network policies:
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: deny-netpol-changes
spec:
validationFailureAction: enforce
background: false
rules:
- name: check-netpol-updates
match:
resources:
kinds:
- NetworkPolicy
name:
- *-default
exclude:
clusterRoles:
- cluster-admin
validate:
message: "Changing default network policies is not allowed"
deny: {}
This policy sets the imagePullPolicy to Always if the image tag is latest:
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: set-image-pull-policy
spec:
rules:
- name: set-image-pull-policy
match:
resources:
kinds:
- Pod
mutate:
overlay:
spec:
containers:
# match images which end with :latest
- (image): "*:latest"
# set the imagePullPolicy to "Always"
imagePullPolicy: "Always"
This policy sets the Zookeeper and Kafka connection strings for all namespaces.
apiVersion: kyverno.io/v1
kind: ClusterPolicy
metadata:
name: "zk-kafka-address"
spec:
rules:
- name: "zk-kafka-address"
match:
resources:
kinds:
- Namespace
generate:
kind: ConfigMap
name: zk-kafka-address
# generate the resource in the new namespace
namespace: "{{request.object.metadata.name}}"
data:
kind: ConfigMap
data:
ZK_ADDRESS: "192.168.10.10:2181,192.168.10.11:2181,192.168.10.12:2181"
KAFKA_ADDRESS: "192.168.10.13:9092,192.168.10.14:9092,192.168.10.15:9092"
For more examples, refer to a list of curated of sample policies that can be applied to your cluster.
- Introducing Kyverno - blog post
- CNCF Video and Slides
- 10 Kubernetes Best Practices - blog post
- VMware Code Meetup Video
- Virtual Rejekts Video
- TGIK Video
Open Policy Agent (OPA) is a general-purpose policy engine that can be used as a Kubernetes admission controller. It supports a large set of use cases. Policies are written using Rego a custom query language.
k-rail provides several ready to use policies for security and multi-tenancy. The policies are written in Golang. Several of the Kyverno sample policies were inspired by k-rail policies.
Polaris validates configurations for best practices. It includes several checks across health, networking, security, etc. Checks can be assigned a severity. A dashboard reports the overall score.
Tools like Kustomize can be used to manage variations in configurations outside of clusters. There are several advantages to this approach when used to produce variations of the same base configuration. However, such solutions cannot be used to validate or enforce configurations.
See Milestones and Issues.
- For feature requests and bugs, file an issue.
- For discussions or questions, join the #kyverno channel on the Kubernetes Slack or the mailing list
Thanks for your interest in contributing!
- Please review and agree to abide with the Code of Conduct before contributing.
- We encourage all contributions and encourage you to read our contribution guidelines.
- See the Wiki for developer documentation.
- Browse through the open issues