Data persistence ================ ```{note} The files for this section are located at . ``` To persist data in the PostgreSQL database in the kanban app (see the previous two sections), we need to define a **persistent volume**. Before we do so, we also need to restart our K8s cluster with a bind mount between our host system and the K8s node: ```{code-block} yaml :caption: db.yaml kind: Cluster apiVersion: kind.x-k8s.io/v1alpha4 nodes: - role: control-plane extraMounts: - hostPath: /home/yinchi/k8s-docs/helm-kanban-v2/mnt/bitnami/postgresql containerPath: /bitnami/postgresql ``` ```{note} We are using the default database location for the PostgreSQL Helm chart for our `containerPath`. This is controlled by the `primary.persistence.mountPath`. If we had a multinode setup, we would also need to configure `readReplicas.persistence.mountPath`. ``` We provide the above configuration when starting up our Kind cluster: ```console $ kind create cluster --config db.yaml Creating cluster "kind" ... ✓ Ensuring node image (kindest/node:v1.30.0) đŸ–ŧ ✓ Preparing nodes đŸ“Ļ ✓ Writing configuration 📜 ✓ Starting control-plane 🕹ī¸ ✓ Installing CNI 🔌 ✓ Installing StorageClass 💾 Set kubectl context to "kind-kind" You can now use your cluster with: kubectl cluster-info --context kind-kind Have a question, bug, or feature request? Let us know! https://kind.sigs.k8s.io/#community 🙂 ``` ## The persistent volume and claim Create the following YAML file: ```{code-block} yaml :caption: pv-volume.yaml :emphasize-lines: 14 apiVersion: v1 kind: PersistentVolume metadata: name: pv-volume labels: type: local spec: storageClassName: "" capacity: storage: 10Gi accessModes: - ReadWriteOnce hostPath: path: "/bitnami/postgresql" claimRef: name: pv-claim namespace: kanban --- apiVersion: v1 kind: PersistentVolumeClaim metadata: name: pv-claim namespace: kanban spec: storageClassName: "" volumeName: pv-volume accessModes: - ReadWriteOnce resources: requests: storage: 3Gi ``` Note that the `claimRef` and `volumeName` above are designed to ensure that the PersistentVolumeClaim and PersistentVolume can only claim/be claimed by each other. Apply the above file to the cluster: ```console $ kubectl create ns kanban namespace/kanban created $ kubectl apply -f pv-volume.yaml -n kanban persistentvolume/pv-volume configured persistentvolumeclaim/pv-claim created ``` ## Modifications to `postgres.yaml` We supply some additional values to `postgres.yaml`: ```{code-block} yaml :caption: postgres.yaml fullnameOverride: "postgres" auth: postgresPassword: "postgres" username: "kanban" password: "kanban" database: "kanban" primary: persistence: existingClaim: "pv-claim" volumePermissions: enabled: true ``` ```console $ helm upgrade postgres bitnami/postgresql -n kanban \ --install --values postgres.yaml --wait --timeout 60s Release "postgres" does not exist. Installing it now. NAME: postgres LAST DEPLOYED: Wed Aug 7 19:35:15 2024 NAMESPACE: kanban STATUS: deployed REVISION: 1 TEST SUITE: None NOTES: ``` The `primary.persistence.existingClaim` value links our database to our mounted directory on disk, while the `volumePermissions.enabled` value (`true`) sets grants read/write permissions for our directory to the PostgreSQL pod (which operates with user ID 1001 by default). We can check this by entering this command on our host machine: ```console $ ls -lah mnt/bitnami/postgresql/ total 12K drwxrwxr-x 3 1001 libvirtd 4.0K Aug 7 19:09 . drwxrwxr-x 3 yinchi yinchi 4.0K Aug 7 18:09 .. drwx------ 19 1001 libvirtd 4.0K Aug 7 19:10 data ``` Launch the backend and frontend as in previous sections: ```console $ helm upgrade kanban-backend ./app -n kanban \ --install --values kanban-backend.yaml $ helm upgrade kanban-frontend ./app -n kanban \ --install --values kanban-frontend.yaml $ kubectl port-forward svc/kanban-frontend 8081:8080 -n kanban Forwarding from 127.0.0.1:8081 -> 80 Forwarding from [::1]:8081 -> 80 ``` Our kanban app should now be available as before. Create a kanban with several tasks. Next, try uninstalling and reinstalling the kanban app, but do not delete the namespace or persistent volume/volume claim: ```console $ helm uninstall kanban-frontend kanban-backend postgres -n kanban release "kanban-frontend" uninstalled release "kanban-backend" uninstalled release "postgres" uninstalled $ helm upgrade postgres bitnami/postgresql -n kanban \ --install --values postgres.yaml --wait --timeout 60s $ helm upgrade kanban-backend ./app -n kanban \ --install --values kanban-backend.yaml $ helm upgrade kanban-frontend ./app -n kanban \ --install --values kanban-frontend.yaml $ kubectl port-forward svc/kanban-frontend 8081:8080 -n kanban Forwarding from 127.0.0.1:8081 -> 80 Forwarding from [::1]:8081 -> 80 ``` Your kanbans should be in the same state as before reinstalling the app. ## Lifecycle of a persistent volume Our PersistentVolume has a policy of `Retain`, meaning its resources are kept even after the PersistentVolume is deleted: ```console $ kubectl get pv NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS VOLUMEATTRIBUTESCLASS REASON AGE pv-volume 10Gi RWO Retain Bound kanban/pv-claim 5m46s ``` We can confirm this by destroying our app and checking the state of the files on disk: ```console $ helm uninstall kanban-frontend kanban-backend postgres -n kanban release "kanban-frontend" uninstalled release "kanban-backend" uninstalled release "postgres" uninstalled $ kubectl delete ns/kanban namespace "kanban" deleted $ kubectl delete pv/pv-volume persistentvolume "pv-volume" deleted $ kind delete cluster Deleting cluster "kind" ... Deleted nodes: ["kind-control-plane"] $ sudo ls mnt/bitnami/postgresql/data/ -1 base global pg_commit_ts pg_dynshmem pg_ident.conf pg_logical pg_multixact pg_notify pg_replslot pg_serial pg_snapshots pg_stat pg_stat_tmp pg_subtrans pg_tblspc pg_twophase PG_VERSION pg_wal pg_xact postgresql.auto.conf postmaster.opts ``` Our database files are still present. Recreate everything we need (output omitted for brevity): ```bash kind create cluster kubectl create ns kanban kubectl apply -f pv-volume.yaml helm upgrade postgres bitnami/postgresql -n kanban \ --create-namespace --install --values postgres.yaml --wait --timeout 60s helm upgrade kanban-backend ./app -n kanban --install --values kanban-backend.yaml helm upgrade kanban-frontend ./app -n kanban --install --values kanban-frontend.yaml kubectl port-forward svc/kanban-frontend 8081:8080 -n kanban ``` We can see our previously created kanbans in the app.