Polyaxon in-cluster NFS provisioner to simplify the creation of ReadWriteMany and ReadOnlyMany volumes.

Overview

This guide shows how to use the NFS provisioner to mount logs to your jobs and experiments.

polyaxon-nfs-provisioner provides a stable Helm chart, maintained and supported by Polyaxon, to easily deploy and spin NFS-volumes to use with Polyaxon.

Install

To install the nfs-provisioner, make sure you have helm installed, please see this guide.

Namespace

If you are using this chart with Polyaxon, please install the chart on the same namespace where you installed Polyaxon.

$ kubectl create namespace polyaxon

namespace "polyaxon" created

Polyaxon's charts repo

You can add the Polyaxon helm repository to your helm, so you can install Polyaxon and other charts provided by Polyaxon from it. This makes it easy to refer to the chart without having to use a long URL each time.

Install the nfs provisioner

helm install polyaxon/nfs-provisioner --name=plxtf --namespace=polyaxon

Configuration

To enable the logs with default values:

logs:
  size: 100Gi
  enabled: true

To enable the logs with custom values:

logs:
  enabled: true

Full reference:

Parameter Description Default
logs.name Name of the PVC to create polyaxon-pvc-logs
logs.size Size of logs volume 10Gi
logs.accessMode Use volume as ReadOnly or ReadWrite ReadWriteOnce ReadWriteMany

Now you can use the PVC to mount logs to your experiments and jobs in Polyaxon

logs:
  existingClaim: polyaxon-pvc-logs
  mountPath: /plx-data