Deploying on Kubernetes

Deploying with Vineyard Operator

Please refer to Vineyard Operator.

Deploying using Python API

For environments where the Helm is not available, vineyard provides a Python API for deploying on Kubernetes,

from vineyard.deploy.kubernetes import start_vineyardd

resources = start_vineyardd()

The deployed etcd pods, vineyard daemonset and services can be viewed as:

$ kubectl -n vineyard get pods
NAME             READY   STATUS    RESTARTS   AGE
etcd0            1/1     Running   0          10s
vineyard-pwkcn   1/1     Running   0          10s

$ kubectl -n vineyard get daemonsets
NAME       DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR   AGE
vineyard   1         1         1       1            1           <none>          14s

$ kubectl -n vineyard get services
NAME                TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
etcd-for-vineyard   ClusterIP   ........        <none>        2379/TCP   18s
vineyard-rpc        ClusterIP   ........        <none>        9600/TCP   18s

Later the cluster can be stoped by releasing the resources with

from vineyard.deploy.kubernetes import delete_kubernetes_objects

delete_kubernetes_objects(resources)

For more details about the API usage, please refer to the Deployment.