Down Provisioner

Explore this Page

• Overview
• DiskPoolClaim CRD
• Benefits of Down Provisioner

Overview

Provisioning storage resources in Kubernetes, especially in cloud environments, can be tedious and error-prone when done manually. The Down Provisioner simplifies this process by automating the creation of Replicated PV Mayastor DiskPools using PersistentVolumeClaims (PVCs) from cloud storage providers that support dynamic provisioning.

With the help of a Kubernetes operator and a custom resource definition (CRD) called DiskPoolClaim, you can declaratively provision DiskPools by referencing existing PVCs and defining your desired configuration. This streamlines storage setup and reduces the need for manual disk discovery or configuration.

DiskPoolClaim CRD

When you install DataCore Puls8, a CRD named DiskPoolClaim is automatically registered in your Kubernetes cluster. You can use this CRD to define a custom resource (CR) that specifies which PVC to use as the backing storage for a DiskPool, along with additional configuration options such as node selection, topology, and encryption.

Create a Basic DiskPool

To create a DiskPool with minimal configuration, define a DiskPoolClaim:

apiVersion: "puls8.datacore.com/v0"
kind: DiskPoolClaim
metadata:
  name: my-first-diskpool
  namespace: puls8
spec:
  nodeName: node-worker-1
  source:
    pvc:
      name: my-cloud-pvc

In this example:

• Replace node-worker-1 with the name of the node where you want to create the DiskPool.
• Replace my-cloud-pvc with the name of a PVC that you have provisioned using your cloud provider.

Make sure your PVC is created with volumeMode: Block and exists in the same namespace where DataCore Puls8 is deployed. Using the WaitForFirstConsumer volume binding mode is recommended to allow the cloud provider to optimize volume placement.

Add Topology Labels

If you need to associate topology labels with the DiskPool for scheduling or placement purposes, you can extend your DiskPoolClaim.

apiVersion: "puls8.datacore.com/v0"
kind: DiskPoolClaim
metadata:
  name: my-first-diskpool
  namespace: puls8
spec:
  nodeName: node-worker-1
  source:
    pvc:
      name: my-cloud-pvc
  topology:
    labelled:
      topology-key: topology-value

Refer to the StorageClass Parameters Documentation for more information.

Enable Encryption

To enable encryption for your DiskPool, include an encryptionConfig section that references a Kubernetes Secret.

apiVersion: "puls8.datacore.com/v0"
kind: DiskPoolClaim
metadata:
  name: my-first-diskpool
  namespace: puls8
spec:
  nodeName: node-worker-1
  source:
    pvc:
      name: my-cloud-pvc
  topology:
    labelled:
      topology-key: topology-value
  encryptionConfig:
    source:
      secret:
        name: pool-secret

Refer to the Encryption Documentation for more information.

Benefits of Down Provisioner

• Automation of DiskPool Creation: You no longer need to manually discover and configure block devices. Down Provisioner automates this process using familiar Kubernetes primitives like PVCs.
• Cloud Integration: You can take full advantage of your cloud provider’s dynamic provisioning capabilities. By referencing cloud-provisioned PVCs, you can align storage creation with your provider’s best practices.
• Declarative Configuration: You define your storage setup using YAML, which simplifies version control, auditing, and repeatability across environments.
• Node-Level Control: You explicitly control where your storage is placed by specifying the target node, which is essential for performance tuning and high availability scenarios
• Support for Topology and Encryption: You can include topology labels for intelligent workload placement and configure encryption to meet data security and compliance requirements.
• Reduced Operational Overhead: With streamlined provisioning and integration into Kubernetes workflows, you can reduce time spent on manual tasks and focus on higher-level operations.

Learn More

• Platform Guides
• Backup and Restore
• DataCore Puls8 Kubectl Plugin