OpenKruise/Kruise
Community Meeting |
---|
The Kruise Project holds bi-weekly community calls. To join us and watch previous meeting notes and recordings, please see meeting schedule. |
Kruise is the core of the OpenKruise project. It is a set of controllers which extends and complements Kubernetes core controllers on workload management.
Today, Kruise offers three workload controllers:
-
Advanced StatefulSet: An enhanced version of default StatefulSet with extra functionalities such as
inplace-update
. -
BroadcastJob: A job that runs Pods to completion across all the nodes in the cluster.
-
SidecarSet: A controller that injects sidecar containers into the Pod spec based on selectors.
Please see documents for more technical information.
Please find the project roadmap here.
Several tutorials are provided to demonstrate how to use the workload controllers.
Getting started
Install with YAML files
Install CRDs
kubectl apply -f https://raw.githubusercontent.com/kruiseio/kruise/master/config/crds/apps_v1alpha1_broadcastjob.yaml
kubectl apply -f https://raw.githubusercontent.com/kruiseio/kruise/master/config/crds/apps_v1alpha1_sidecarset.yaml
kubectl apply -f https://raw.githubusercontent.com/kruiseio/kruise/master/config/crds/apps_v1alpha1_statefulset.yaml
Note that ALL three CRDs need to be installed for kruise-controller to run properly.
Install kruise-controller-manager
kubectl apply -f https://raw.githubusercontent.com/kruiseio/kruise/master/config/manager/all_in_one.yaml
Or run from the repo root directory:
kustomize build config/default | kubectl apply -f -
To Install Kustomize, check kustomize website.
Note that use Kustomize 1.0.11. Version 2.0.3 has compatibility issues with kube-builder
The official kruise-controller-manager image is hosted under docker hub.
Usage examples
Advanced StatefulSet
apiVersion: apps.kruise.io/v1alpha1
kind: StatefulSet
metadata:
name: sample
spec:
replicas: 3
serviceName: fake-service
selector:
matchLabels:
app: sample
template:
metadata:
labels:
app: sample
spec:
readinessGates:
# A new condition must be added to ensure the pod remain at NotReady state while the in-place update is happening
- conditionType: InPlaceUpdateReady
containers:
- name: main
image: nginx:alpine
podManagementPolicy: Parallel # allow parallel updates, works together with maxUnavailable
updateStrategy:
type: RollingUpdate
rollingUpdate:
# Do in-place update if possible, currently only image update is supported for in-place update
podUpdatePolicy: InPlaceIfPossible
# Allow parallel updates with max number of unavailable instances equals to 2
maxUnavailable: 2
Broadcast Job
Run a BroadcastJob that each Pod computes pi, with ttlSecondsAfterFinished
set to 30. The job
will be deleted in 30 seconds after the job is finished.
apiVersion: apps.kruise.io/v1alpha1
kind: BroadcastJob
metadata:
name: broadcastjob-ttl
spec:
template:
spec:
containers:
- name: pi
image: perl
command: ["perl", "-Mbignum=bpi", "-wle", "print bpi(2000)"]
restartPolicy: Never
completionPolicy:
type: Always
ttlSecondsAfterFinished: 30
SidecarSet
The yaml file below describes a SidecarSet that contains a sidecar container named sidecar1
# sidecarset.yaml
apiVersion: apps.kruise.io/v1alpha1
kind: SidecarSet
metadata:
name: test-sidecarset
spec:
selector: # select the pods to be injected with sidecar containers
matchLabels:
app: nginx
containers:
- name: sidecar1
image: centos:7
command: ["sleep", "999d"] # do nothing at all
Developer Guide
There's a Makefile
in the root folder which describes the options to build and install. Here are some common ones:
Build the controller manager binary
make manager
Run the tests
make test
Build the docker image, by default the image name is openkruise/kruise-manager:v1alpha1
export IMG=<your_image_name> && make docker-build
Push the image
export IMG=<your_image_name> && make docker-push
or just
docker push <your_image_name>
Generate manifests e.g. CRD, RBAC etc.
make manifests
Community
If you have any questions or want to contribute, you are welcome to communicate most things via GitHub issues or pull requests.
Other active communication channels:
- Slack: channel address
- Mailing List: todo
- Dingtalk Group(钉钉讨论群)
Copyright
Certain implementation relies on existing code from Kubernetes and the credit goes to original Kubernetes authors.