You are looking at the documentation of a prior release. To read the documentation of the latest release, please visit here.

New to Stash? Please start here.

Using Stash with Google Kubernetes Engine (GKE)

This tutorial will show you how to use Stash to backup and restore a Kubernetes deployment in Google Kubernetes Engine. Here, we are going to back up the /source/data folder of a busybox pod into GCS bucket. Then, we are going to show how to recover this data into a gcePersistentDisk and PersistentVolumeClaim. We are going to also re-deploy deployment using this recovered volume.

Before You Begin

At first, you need to have a Kubernetes cluster in Google Cloud Platform. If you don’t already have a cluster, you can create one from here.

To keep things isolated, we are going to use a separate namespace called demo throughout this tutorial.

$ kubectl create ns demo
namespace/demo created

Note: YAML files used in this tutorial are stored in /docs/examples/platforms/gke directory of stashed/docs repository.

Backup

In order to take backup, we need some sample data. Stash has some sample data in stash-data repository. As gitRepo volume has been deprecated, we are not going to use this repository as volume directly. Instead, we are going to create a configMap from the stash-data repository and use that ConfigMap as data source.

Let’s create a ConfigMap from these sample data,

$ kubectl create configmap -n demo stash-sample-data \
	--from-literal=LICENSE="$(curl -fsSL https://github.com/stashed/stash-data/raw/master/LICENSE)" \
	--from-literal=README.md="$(curl -fsSL https://github.com/stashed/stash-data/raw/master/README.md)"
configmap/stash-sample-data created

Deploy Workload:

Now, deploy the following Deployment. Here, we have mounted the ConfigMap stash-sample-data as data source volume.

Below, the YAML for the Deployment we are going to create.

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: stash-demo
  name: stash-demo
  namespace: demo
spec:
  replicas: 1
  selector:
    matchLabels:
      app: stash-demo
  template:
    metadata:
      labels:
        app: stash-demo
      name: busybox
    spec:
      containers:
      - args:
        - sleep
        - "3600"
        image: busybox
        imagePullPolicy: IfNotPresent
        name: busybox
        volumeMounts:
        - mountPath: /source/data
          name: source-data
      restartPolicy: Always
      volumes:
      - name: source-data
        configMap:
          name: stash-sample-data

Let’s create the deployment we have shown above,

$ kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/deployment.yaml
deployment.apps/stash-demo created

Now, wait for deployment’s pod to go in Running state.

$ kubectl get pods -n demo -l app=stash-demo
NAME                         READY     STATUS    RESTARTS   AGE
stash-demo-b66b9cdfd-8s98d   1/1       Running   0          6m

You can check that the /source/data/ directory of pod is populated with data from the volume source using this command,

$ kubectl exec -n demo stash-demo-b66b9cdfd-8s98d -- ls -R /source/data/
/source/data:
LICENSE
README.md

Now, we are ready backup /source/data directory into a GCS bucket,

Create Secret:

At first, we need to create a storage secret that hold the credentials for the backend. To configure this backend, the following secret keys are needed:

KeyDescription
RESTIC_PASSWORDRequired. Password used to encrypt snapshots by restic
GOOGLE_PROJECT_IDRequired. Google Cloud project ID
GOOGLE_SERVICE_ACCOUNT_JSON_KEYRequired. Google Cloud service account json key

Create storage secret as below,

$ echo -n 'changeit' > RESTIC_PASSWORD
$ echo -n '<your-project-id>' > GOOGLE_PROJECT_ID
$ cat /path/to/downloaded-sa-json.key > GOOGLE_SERVICE_ACCOUNT_JSON_KEY
$ kubectl create secret generic -n demo gcs-secret \
    --from-file=./RESTIC_PASSWORD \
    --from-file=./GOOGLE_PROJECT_ID \
    --from-file=./GOOGLE_SERVICE_ACCOUNT_JSON_KEY
secret "gcs-secret" created

Verify that the secret has been created successfully,

$ kubectl get secret -n demo gcs-secret -o yaml
apiVersion: v1
data:
  GOOGLE_PROJECT_ID: <base64 encoded google project id>
  GOOGLE_SERVICE_ACCOUNT_JSON_KEY: <base64 encoded google service account json key>
  RESTIC_PASSWORD: Y2hhbmdlaXQ=
kind: Secret
metadata:
  creationTimestamp: 2018-04-11T12:57:05Z
  name: gcs-secret
  namespace: demo
  resourceVersion: "7113"
  selfLink: /api/v1/namespaces/demo/secrets/gcs-secret
  uid: d5e70521-3d87-11e8-a5b9-42010a800002
type: Opaque

Create Restic:

Now, we can create Restic crd. This will create a repository in the GCS bucket specified in gcs.bucket field and start taking periodic backup of /source/data directory.

$ kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/restic.yaml
restic.stash.appscode.com/gcs-restic created

Below, the YAML for Restic crd we have created above,

apiVersion: stash.appscode.com/v1alpha1
kind: Restic
metadata:
  name: gcs-restic
  namespace: demo
spec:
  selector:
    matchLabels:
      app: stash-demo
  fileGroups:
  - path: /source/data
    retentionPolicyName: 'keep-last-5'
  backend:
    gcs:
      bucket: stash-backup-repo
      prefix: demo
    storageSecretName: gcs-secret
  schedule: '@every 1m'
  volumeMounts:
  - mountPath: /source/data
    name: source-data
  retentionPolicies:
  - name: 'keep-last-5'
    keepLast: 5
    prune: true

If everything goes well, Stash will inject a sidecar container into the stash-demo deployment to take periodic backup. Let’s check sidecar has been injected successfully,

$ kubectl get pod -n demo -l app=stash-demo
NAME                          READY   STATUS    RESTARTS   AGE
stash-demo-6b8c94cdd7-8jhtn   2/2     Running   1          1h

Look at the pod. It now has 2 containers. If you view the resource definition of this pod, you will see there is a container named stash which running backup command.

Verify Backup:

Stash will create a Repository crd with name deployment.stash-demo for the respective repository in GCS backend. To verify, run the following command,

$ kubectl get repository deployment.stash-demo -n demo
NAME                    BACKUPCOUNT   LASTSUCCESSFULBACKUP   AGE
deployment.stash-demo   1             13s                    1m

Here, BACKUPCOUNT field indicates number of backup snapshot has taken in this repository.

Restic will take backup of the volume periodically with a 1-minute interval. You can verify that backup is taking successfully by,

$ kubectl get snapshots -n demo -l repository=deployment.stash-demo
NAME                             AGE
deployment.stash-demo-c1014ca6   10s

Here, deployment.stash-demo-c1014ca6 represents the name of the successful backup Snapshot taken by Stash in deployment.stash-demo repository.

If you navigate to <bucket name>/demo/deployment/stash-demo directory in your GCS bucket. You will see, a repository has been created there.

  Repository in GCS Bucket

To view the snapshot files, navigate to snapshots directory of the repository,

  Snapshot in GCS Bucket

Stash keeps all backup data encrypted. So, snapshot files in the bucket will not contain any meaningful data until they are decrypted.

Recovery

Now, consider that we have lost our workload as well as data volume. We want to recover the data into a new volume and re-deploy the workload. In this section, we are going to see how to recover data into a gcePersistentDisk and persistentVolumeClaim.

At first, let’s delete Restic crd, stash-demo deployment and stash-sample-data ConfigMap.

$ kubectl delete deployment -n demo stash-demo
deployment.extensions "stash-demo" deleted

$ kubectl delete restic -n demo gcs-restic
restic.stash.appscode.com "gcs-restic" deleted

$ kubectl delete configmap -n demo stash-sample-data
configmap "stash-sample-data" deleted

In order to perform recovery, we need Repository crd deployment.stah-demo and backend secret gcs-secret to exist.

In case of cluster disaster, you might lose Repository crd and backend secret. In this scenario, you have to create the secret again and Repository crd manually. Follow the guide to understand Repository crd structure from here.

Recover to GCE Persistent Disk

Now, we are going to recover the backed up data into GCE Persistent Disk. At first, create a GCE disk named stash-recovered from Google cloud console. Then create Recovery crd,

$ kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/recovery-gcePD.yaml
recovery.stash.appscode.com/gcs-recovery created

Below, the YAML for Recovery crd we have created above.

apiVersion: stash.appscode.com/v1alpha1
kind: Recovery
metadata:
  name: gcs-recovery
  namespace: demo
spec:
  repository:
    name: deployment.stash-demo
    namespace: demo
  paths:
  - /source/data
  recoveredVolumes:
  - mountPath: /source/data
    gcePersistentDisk:
        pdName: stash-recovered
        fsType: ext4

Wait until Recovery job completes its task. To verify that recovery has completed successfully run,

$ kubectl get recovery -n demo gcs-recovery
NAME             REPOSITORYNAMESPACE   REPOSITORYNAME          SNAPSHOT   PHASE       AGE
gcs-recovery     demo                  deployment.stash-demo              Succeeded   3m

Here, PHASE Succeeded indicate that our recovery has been completed successfully. Backup data has been restored in stash-recovered Persistent Disk. Now, we are ready to use this Persistent Disk to re-deploy workload.

If you are using Kubernetes version older than v1.11.0 then run following command and check status.phase field to see whether the recovery succeeded or failed.

$ kubectl get recovery -n demo gcs-recovery -o yaml

Re-deploy Workload:

We have successfully restored backup data into stash-recovered gcePersistentDisk. Now, we are going to re-deploy our previous deployment stash-demo. This time, we are going to mount the stash-recovered Persistent Disk as source-data volume instead of ConfigMap stash-sample-data.

Below, the YAML for stash-demo deployment with stash-recovered persistent disk as source-data volume.

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: stash-demo
  name: stash-demo
  namespace: demo
spec:
  replicas: 1
  selector:
    matchLabels:
      app: stash-demo
  template:
    metadata:
      labels:
        app: stash-demo
      name: busybox
    spec:
      containers:
      - args:
        - sleep
        - "3600"
        image: busybox
        imagePullPolicy: IfNotPresent
        name: busybox
        volumeMounts:
        - mountPath: /source/data
          name: source-data
      restartPolicy: Always
      volumes:
      - name: source-data
        gcePersistentDisk:
          pdName: stash-recovered
          fsType: ext4

Let’s create the deployment,

$  kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/restored-deployment-gcePD.yaml
deployment.apps/stash-demo created

Verify Recovered Data:

We have re-deployed stash-demo deployment with recovered volume. Now, it is time to verify that the data are present in /source/data directory.

Get the pod of new deployment,

$ kubectl get pods -n demo -l app=stash-demo
NAME                         READY     STATUS    RESTARTS   AGE
stash-demo-857995799-gpml9   1/1       Running   0          34s

Run following command to view data of /source/data directory of this pod,

$ kubectl exec -n demo stash-demo-857995799-gpml9 -- ls -R /source/data
/source/data:
LICENSE
README.md
lost+found

/source/data/lost+found:

So, we can see that the data we had backed up from original deployment are now present in re-deployed deployment.

Recover to PersistentVolumeClaim

Here, we are going to show how to recover the backed up data into a PVC. If you have re-deployed stash-demo deployment by following previous tutorial on gcePersistentDisk, delete the deployment first,

$ kubectl delete deployment -n demo stash-demo
deployment.apps/stash-demo deleted

Now, create a PersistentVolumeClaim where our recovered data will be stored.

$ kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/pvc.yaml
persistentvolumeclaim/stash-recovered created

Below the YAML for PersistentVolumeClaim we have created above,

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: stash-recovered
  namespace: demo
  labels:
    app: stash-demo
spec:
  storageClassName: standard
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 2Gi

Check that if cluster has provisioned the requested claim,

$ kubectl get pvc -n demo -l app=stash-demo
NAME              STATUS    VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS   AGE
stash-recovered   Bound     pvc-57bec6e5-3e11-11e8-951b-42010a80002e   2Gi        RWO            standard       1m

Look at the STATUS filed. stash-recovered PVC is bounded to volume pvc-57bec6e5-3e11-11e8-951b-42010a80002e.

Create Recovery:

Now, we have to create a Recovery crd to recover backed up data into this PVC.

$ kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/recovery-pvc.yaml
recovery.stash.appscode.com/gcs-recovery created

Below, the YAML for Recovery crd we have created above.

apiVersion: stash.appscode.com/v1alpha1
kind: Recovery
metadata:
  name: gcs-recovery
  namespace: demo
spec:
  repository:
    name: deployment.stash-demo
    namespace: demo
  paths:
  - /source/data
  recoveredVolumes:
  - mountPath: /source/data
    persistentVolumeClaim:
      claimName: stash-recovered

Wait until Recovery job completes its task. To verify that recovery has completed successfully run,

$ kubectl get recovery -n demo gcs-recovery
NAME             REPOSITORYNAMESPACE   REPOSITORYNAME          SNAPSHOT   PHASE       AGE
gcs-recovery     demo                  deployment.stash-demo              Succeeded   3m

Here, PHASE Succeeded indicate that our recovery has been completed successfully. Backup data has been restored in stash-recovered PVC. Now, we are ready to use this PVC to re-deploy workload.

Re-deploy Workload:

We have successfully restored backup data into stash-recovered PVC. Now, we are going to re-deploy our previous deployment stash-demo. This time, we are going to mount the stash-recovered PVC as source-data volume instead of ConfigMap stash-sample-data.

Below, the YAML for stash-demo deployment with stash-recovered PVC as source-data volume.

apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: stash-demo
  name: stash-demo
  namespace: demo
spec:
  replicas: 1
  selector:
    matchLabels:
      app: stash-demo
  template:
    metadata:
      labels:
        app: stash-demo
      name: busybox
    spec:
      containers:
      - args:
        - sleep
        - "3600"
        image: busybox
        imagePullPolicy: IfNotPresent
        name: busybox
        volumeMounts:
        - mountPath: /source/data
          name: source-data
      restartPolicy: Always
      volumes:
      - name: source-data
        persistentVolumeClaim:
          claimName: stash-recovered

Let’s create the deployment,

$ kubectl apply -f https://github.com/stashed/docs/raw/v2020.11.17/docs/examples/platforms/gke/restored-deployment-pvc.yaml
deployment.apps/stash-demo created

Verify Recovered Data:

We have re-deployed stash-demo deployment with recovered volume. Now, it is time to verify that the data are present in /source/data directory.

Get the pod of new deployment,

$ kubectl get pods -n demo -l app=stash-demo
NAME                          READY     STATUS    RESTARTS   AGE
stash-demo-559845c5db-8cd4w   1/1       Running   0          33s

Run following command to view data of /source/data directory of this pod,

$ kubectl exec -n demo stash-demo-559845c5db-8cd4w -- ls -R /source/data
/source/data:
LICENSE
README.md
lost+found

/source/data/lost+found:

So, we can see that the data we had backed up from original deployment are now present in re-deployed deployment.

Cleanup

To cleanup the resources created by this tutorial, run following commands:

$ kubectl delete recovery -n demo gcs-recovery
$ kubectl delete secret -n demo gcs-secret
$ kubectl delete deployment -n demo stash-demo
$ kubectl delete pvc -n demo stash-recovered
$ kubectl delete repository -n demo deployment.stash-demo
$ kubectl delete ns demo
  • To uninstall Stash from your cluster, follow the instructions from here.