Persistence Storage Volume for GCP k8s

[coolpalani]

I have deployed the K8s on GCP. How do we provision persistence storage or Loadbalancer from GCP. Could you provide some details. I didn't see any proper documents. Even In kubernetes.io site also, they are providing details only for AWS, OpenStack.

In this below link, they had explained detail for OpenStack.

https://medium.com/@bugosip/kubeadm-openstack-cloud-provider-using-kubernetes-1-9-24882f43fab6

I am looking something similar to GCP. :)

Thanks in advance.

[ageapps]

Sorry it took me so long to answer.

Since k8s version 1.8 things related to this have changed. Looking at the documentation (https://kubernetes.io/docs/concepts/storage/storage-classes/) the storage privisioning is done using the StorageClass resource inside K8s where you can specify several parameters like the provisioner (in your case 'GCEPersistentDisk'). Another thing you should look are the Storage Claims inside k8s (https://kubernetes.io/docs/concepts/storage/persistent-volumes).

Regarding the LoadBalancers, as you know, there are several ways to expose a k8s service outside of your cluster, as far as i know, there are 2 that provision a LoadBalancer inside Google's cloud plattform. In both cases k8s internally uses its integration with the corresponding cloud provider in order to get the resources needed:

• When setting the service type to LoadBalancer (https://kubernetes.io/docs/concepts/services-networking/service/). Here google just gives you a public ip assigned to that service
• When defining an ingress(https://kubernetes.io/docs/concepts/services-networking/ingress/) resource inside your cluster. In this case, the ingress is managed by the ingress controller in your cluster (Google deploys it for you inside the kube-system namespace) and it provisions an external LoadBalancer managed by Google.

In any case, it's important to note that K8s is a solution that is usually deployed on top of a cloud infrastructure and the main Cloud Providers have their own products integrated inside the k8s ecosystem making posible to manage some resources like storage, ingresses, security, etc from inside your cluster.

I hope it helps in any way