Deploy a SingleInstance Database with a static NFS filesystem

Introduction

Static Persistence

In this lab we'll be using one of the Static Persistence options for our database, so we'll create manually a persistent volume and refer to it in our database configuration file.

You could use a block volume for this, but the result would be like in the first lab: this volume can only be mounted on a single node, so no failover options.

Therefore we'll be using an NFS volume, which can be mounted on all 3 nodes at once and allows us to demonstrate what happens when a node goes down.

Assumption : we assume you ran the first lab, so you have created the various secrets required to pull down and spin up a database container.

Estimated Time: 20 minutes

Objective

Create a database running on Kubernetes, using an NFS volume as persistency store

Prerequisites

You have executed Lab 2: Install the Oracle Database Kubernetes Operator

Task 1: Prepare your NFS persistent volume

On the OCI console, navigate to Storage, section File Storage, and select Mount Targets Create a new mount target using the Create Mount Target button.
- Make sure to select the Virtual Cloud Network that was created as part of the OKE Cluster creation - it will have a name starting with oke-vcn-quick-...
- Select the Subnet where the kubernetes nodes are located. This subnet has a name starting with oke-nodesubnet-quick-...
- Leave the other parameters on their default values
You can now hit the Create button.
- After the creation you are redirected to the Details screen of the mount point.
- Note down the IP Address of the mount point, you will need this information later in the setup. In the example above, the address is 10.0.0.4
- Before we move on, you need to change the Reported Size of the mount target. To do this, select the tab NFS to display the NFS parameters of the moint point:
- Now use the pencil icon on the side of the Reported Size parameter:
- After clicking the pencil, enter a value of 2000 GiB in the edit box, and hit the Save icon
Now navigate to the File Storage, then File Systems menu, and click the Create File System button
- Keep the default workflow: File System for NFS
- All parameters are pre-filled, just validate the selected Mount Target is indeed the mount target you just created
- Take note of the Export Path : you will need this information later in the database configuration file. In the below example the path is /FileSystem-20220727-1600-55
- Click the Create button to finish the operation
- The Utilization of the volume is currently 0 B: the volume has not yet been used. We'll see how this parameter will increase once the DB is created.
- Copy and store the OCID of the filesystem (you can use the Copy button), you will need this info later.
To finalize the configuration of the volume we need to open the appropriate ports on the subnet used to allow the Kubernetes nodes and the NFS volume to communicate with each other :
- Navigate to Networking, and select Virtual Cloud Networks
- Select the VCN that was created for the Kubernetes cluster (name staring with oke-vcn-quick-...)
- Select the Subnet used for the kubernetes nodes (name starting with oke-nodesubnet-quick-...)
- Select the security list of the subnet, with a name starting with oke-nodeseclist-...
- Click the Add Ingress Rule button and create a Stateful ingress from ALL ports in source CIDR block to TCP ports 111, 2048, 2049, and 2050 :
  - Source CIDR : 0.0.0.0/0
  - IP Protocol: TCP
  - Destination Port Range: 111,2048,2049,2050
- Repeat the operation but this time specifying the UDP protocol to set a Stateful ingress from ALL ports in source CIDR block to UDP ports 111 and 2048:
  - Source CIDR : 0.0.0.0/0
  - IP Protocol: UDP
  - Destination Port Range: 111,2048
- Now select the Egress Rules resources on the left of the screen
- Here click the Add Egress Rules to create a Stateful egress from TCP ALL ports to ports 111, 2048, 2049, and 2050 in destination CIDR block.
  - Destination CIDR : 0.0.0.0/0
  - IP Protocol: TCP
  - Destination Port Range: 111,2048,2049,2050
- Repeat the operation but this time specifying the UDP protocol to set a Stateful egress from UDP ALL ports to port 111 in destination CIDR block.
  - Destination CIDR : 0.0.0.0/0
  - IP Protocol: UDP
  - Destination Port Range: 111
Next we need to create a kubernetes persistent volume that points to the configuration we just created. To do this, we'll use the OCI Cloud shell command interface
- Edit a new file to contain the definition of the persistent volume, using the vi or nano editor
  - File name : pv.yaml
  - Paste the below content into the file
```
apiVersion: v1
kind: PersistentVolume
metadata:
  name: nfs-vol-1
spec:
  storageClassName: oci-fss
  capacity:
    storage: 224Gi
  volumeMode: Filesystem
  accessModes:
   - ReadWriteMany
  persistentVolumeReclaimPolicy: Retain
  csi:
    driver: fss.csi.oraclecloud.com
    volumeHandle: "<OCID of the file system>:<Mount Target IP Address>:/<Export Path>"
```
- Replace the placeholders for the parameter volumeHandle:
  - <OCID of the file system> : this is the OCID you noted down after the creation of the filesystem
  - <Mount Target IP Address> : the IP address of the mount point you noted earlier
  - <Export Path> : the export path you noted earlier
  - Example volumeHandle:
```
volumeHandle: "ocid1.filesystem.oc1.eu_frankfurt_1.aaaaaqe3bj...eaaa:10.0.10.156:/FileSystem-20220713-1036-02"
```
- Now apply the config using kubectl:
```
kubectl apply -f pv.yaml
```
  This should show you : persistentvolume/nfs-vol-1 created