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

Backup Elasticsearch using Stash Auto-Backup

Stash can be configured to automatically backup any Elasticsearch database in your cluster. Stash enables cluster administrators to deploy backup blueprints ahead of time so that the database owners can easily backup their database with just a few annotations.

In this tutorial, we are going to show how you can configure a backup blueprint for Elasticsearch databases in your cluster and backup them with few annotations.

Before You Begin

  • At first, you need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster.
  • Install Stash in your cluster following the steps here.
  • Install KubeDB in your cluster following the steps here. This step is optional. You can deploy your database using any method you want.
  • If you are not familiar with how Stash backup and restore Elasticsearch databases, please check the following guide here.
  • If you are not familiar with how auto-backup works in Stash, please check the following guide here.
  • If you are not familiar with the available auto-backup options for databases in Stash, please check the following guide here.

You should be familiar with the following Stash concepts:

In this tutorial, we are going to show backup of three different Elasticsearch databases on three different namespaces named demo, demo-2, and demo-3. Create the namespaces as below if you haven’t done it already.

❯ kubectl create ns demo
namespace/demo created

❯ kubectl create ns demo-2
namespace/demo-2 created

❯ kubectl create ns demo-3
namespace/demo-3 created

Make sure you have installed the Elasticsearch addon for Stash. If you haven’t installed it already, please install the addon following the steps here.

❯ kubectl get tasks.stash.appscode.com | grep elasticsearch
elasticsearch-backup-5.6.4-v6    4d4h
elasticsearch-backup-6.2.4-v6    4d4h
elasticsearch-backup-6.3.0-v6    4d4h
elasticsearch-backup-6.4.0-v6    4d4h
elasticsearch-backup-6.5.3-v6    4d4h
elasticsearch-backup-6.8.0-v6    4d4h
elasticsearch-backup-7.2.0-v6    4d4h
elasticsearch-backup-7.3.2-v6    4d4h
elasticsearch-restore-5.6.4-v6   4d4h
elasticsearch-restore-6.2.4-v6   4d4h
elasticsearch-restore-6.3.0-v6   4d4h
elasticsearch-restore-6.4.0-v6   4d4h
elasticsearch-restore-6.5.3-v6   4d4h
elasticsearch-restore-6.8.0-v6   4d4h
elasticsearch-restore-7.2.0-v6   4d4h
elasticsearch-restore-7.3.2-v6   4d4h

Prepare Backup Blueprint

To backup an Elasticsearch database using Stash, you have to create a Secret containing the backend credentials, a Repository containing the backend information, and a BackupConfiguration containing the schedule and target information. A BackupBlueprint allows you to specify a template for the Repository and the BackupConfiguration.

The BackupBlueprint is a non-namespaced CRD. So, once you have created a BackupBlueprint, you can use it to backup any Elasticsearch database of any namespace just by creating the storage Secret in that namespace and adding few annotations to your Elasticsearch CRO. Then, Stash will automatically create a Repository and a BackupConfiguration according to the template to backup the database.

Below is the BackupBlueprint object that we are going to use in this tutorial,

apiVersion: stash.appscode.com/v1beta1
kind: BackupBlueprint
metadata:
  name: elasticsearch-backup-template
spec:
  # ============== Blueprint for Repository ==========================
  backend:
    gcs:
      bucket: stash-testing
      prefix: stash-backup/${TARGET_NAMESPACE}/${TARGET_APP_RESOURCE}/${TARGET_NAME}
    storageSecretName: gcs-secret
  # ============== Blueprint for BackupConfiguration =================
  task:
    name: elasticsearch-backup-6.3.0-v7
  schedule: "*/5 * * * *"
  interimVolumeTemplate:
    metadata:
      name: ${TARGET_APP_RESOURCE}-${TARGET_NAME} # To ensure that the PVC names are unique for different database
    spec:
      accessModes: [ "ReadWriteOnce" ]
      storageClassName: "standard"
      resources:
        requests:
          storage: 1Gi
  retentionPolicy:
    name: 'keep-last-5'
    keepLast: 5
    prune: true

Here, we are using a GCS bucket as our backend. We are providing gcs-secret at the storageSecretName field. Hence, we have to create a secret named gcs-secret with the access credentials of our bucket in every namespace where we want to enable backup through this blueprint.

Notice the prefix field of backend section. We have used some variables in form of ${VARIABLE_NAME}. Stash will automatically resolve those variables from the database information to make the backend prefix unique for each database instance.

We have also used some variables in name field of the interimVolumeTemplate section. This is to ensure that the generated PVC name becomes unique for the database instances.

Let’s create the BackupBlueprint we have shown above,

❯ kubectl apply -f https://github.com/stashed/elasticsearch/raw/6.3.0-v7/docs/auto-backup/examples/backupblueprint.yaml
backupblueprint.stash.appscode.com/elasticsearch-backup-template created

Now, we are ready to backup our Elasticsearch databases using few annotations. You can check available auto-backup annotations for a databases from here.

Auto-backup with default configurations

In this section, we are going to backup an Elasticsearch database of demo namespace. We are going to use the default configurations specified in the BackupBlueprint.

Create Storage Secret

At first, let’s create the gcs-secret in demo namespace with the access credentials to our GCS bucket.

echo -n 'changeit' > RESTIC_PASSWORD
echo -n '<your-project-id>' > GOOGLE_PROJECT_ID
❯ cat 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

Create Database

Now, we are going to create an Elasticsearch CRO in demo namespace. Below is the YAML of the Elasticsearch object that we are going to create,

apiVersion: kubedb.com/v1alpha2
kind: Elasticsearch
metadata:
  name: es-demo
  namespace: demo
  annotations:
    stash.appscode.com/backup-blueprint: elasticsearch-backup-template
spec:
  version: 7.9.1-xpack
  replicas: 1
  storageType: Durable
  storage:
    resources:
      requests:
        storage: 1Gi
    storageClassName: standard
  terminationPolicy: WipeOut

Notice the annotations section. We are pointing to the BackupBlueprint that we have created earlier though stash.appscode.com/backup-blueprint annotation. Stash will watch this annotation and create a Repository and a BackupConfiguration according to the BackupBlueprint.

Let’s create the above Elasticsearch CRO,

❯ kubectl apply -f https://github.com/stashed/elasticsearch/raw/6.3.0-v7/docs/auto-backup/examples/es-demo.yaml
elasticsearch.kubedb.com/sample-elasticsearch created

Verify Auto-backup configured

In this section, we are going to verify whether Stash has created the respective Repository and BackupConfiguration for our Elasticsearch database we have just deployed or not.

Verify Repository

At first, let’s verify whether Stash has created a Repository for our Elasticsearch or not.

❯ kubectl get repository -n demo
NAME          INTEGRITY   SIZE   SNAPSHOT-COUNT   LAST-SUCCESSFUL-BACKUP   AGE
app-es-demo                                                                5s

Now, let’s check the YAML of the Repository.

❯ kubectl get repository -n demo app-es-demo -o yaml
apiVersion: stash.appscode.com/v1alpha1
kind: Repository
metadata:
...
  name: app-es-demo
  namespace: demo
spec:
  backend:
    gcs:
      bucket: stash-testing
      prefix: stash-backup/demo/elasticsearch/es-demo
    storageSecretName: gcs-secret

Here, you can see that Stash has resolved the variables in prefix field and substituted them with the equivalent information from this database.

Verify BackupConfiguration

Now, let’s verify whether Stash has created a BackupConfiguration for our Elasticsearch or not.

❯ kubectl get backupconfiguration -n demo
NAME          TASK                            SCHEDULE      PAUSED   AGE
app-es-demo   elasticsearch-backup-6.3.0-v7   */5 * * * *            12s

Now, let’s check the YAML of the BackupConfiguration.

❯ kubectl get backupconfiguration -n demo app-es-demo -o yaml
apiVersion: stash.appscode.com/v1beta1
kind: BackupConfiguration
metadata:
  name: app-es-demo
  namespace: demo
  ...
spec:
  driver: Restic
  interimVolumeTemplate:
    metadata:
      name: elasticsearch-es-demo
    spec:
      accessModes:
      - ReadWriteOnce
      resources:
        requests:
          storage: 1Gi
      storageClassName: standard
    status: {}
  repository:
    name: app-es-demo
  retentionPolicy:
    keepLast: 5
    name: keep-last-5
    prune: true
  runtimeSettings: {}
  schedule: '*/5 * * * *'
  target:
    ref:
      apiVersion: appcatalog.appscode.com/v1alpha1
      kind: AppBinding
      name: es-demo
  task:
    name: elasticsearch-backup-6.3.0-v7
  tempDir: {}
status:
  conditions:
  - lastTransitionTime: "2021-02-12T11:46:53Z"
    message: Repository demo/app-es-demo exist.
    reason: RepositoryAvailable
    status: "True"
    type: RepositoryFound
  - lastTransitionTime: "2021-02-12T11:46:53Z"
    message: Backend Secret demo/gcs-secret exist.
    reason: BackendSecretAvailable
    status: "True"
    type: BackendSecretFound
  - lastTransitionTime: "2021-02-12T11:46:53Z"
    message: Backup target appcatalog.appscode.com/v1alpha1 appbinding/es-demo found.
    reason: TargetAvailable
    status: "True"
    type: BackupTargetFound
  - lastTransitionTime: "2021-02-12T11:46:53Z"
    message: Successfully created backup triggering CronJob.
    reason: CronJobCreationSucceeded
    status: "True"
    type: CronJobCreated
  observedGeneration: 1

Notice the interimVolumeTemplate section. The variables of name field have been substituted by the equivalent information from the database.

Also, notice the target section. Stash has automatically added the Elasticsearch as the target of this BackupConfiguration.

Verify Backup

Now, let’s wait for a backup run to complete. You can watch for BackupSession as below,

❯ kubectl get backupsession -n demo -w
NAME                     INVOKER-TYPE          INVOKER-NAME   PHASE       AGE
app-es-demo-1613130605   BackupConfiguration   app-es-demo                0s
app-es-demo-1613130605   BackupConfiguration   app-es-demo    Running     10s
app-es-demo-1613130605   BackupConfiguration   app-es-demo    Succeeded   46s

Once the backup has been completed successfully, you should see the backed up data has been stored in the bucket at the directory pointed by the prefix field of the Repository.

Backup data in GCS Bucket
Fig: Backup data in GCS Bucket

Auto-backup with a custom schedule

In this section, we are going to backup an Elasticsearch database of demo-2 namespace. This time, we are going to overwrite the default schedule used in the BackupBlueprint.

Create Storage Secret

At first, let’s create the gcs-secret in demo-2 namespace with the access credentials to our GCS bucket.

❯ kubectl create secret generic -n demo-2 gcs-secret \
    --from-file=./RESTIC_PASSWORD \
    --from-file=./GOOGLE_PROJECT_ID \
    --from-file=./GOOGLE_SERVICE_ACCOUNT_JSON_KEY
secret/gcs-secret created

Create Database

Now, we are going to create an Elasticsearch CRO in demo-2 namespace. Below is the YAML of the Elasticsearch object that we are going to create,

apiVersion: kubedb.com/v1alpha2
kind: Elasticsearch
metadata:
  name: es-demo-2
  namespace: demo-2
  annotations:
    stash.appscode.com/backup-blueprint: elasticsearch-backup-template
    stash.appscode.com/schedule: "*/3 * * * *"
spec:
  version: 7.9.1-xpack
  replicas: 1
  storageType: Durable
  storage:
    resources:
      requests:
        storage: 1Gi
    storageClassName: standard
  terminationPolicy: WipeOut

Notice the annotations section. This time, we have passed a schedule via stash.appscode.com/schedule annotation along with the stash.appscode.com/backup-blueprint annotation.

Let’s create the above Elasticsearch CRO,

❯ kubectl apply -f https://github.com/stashed/elasticsearch/raw/6.3.0-v7/docs/auto-backup/examples/es-demo-2.yaml
elasticsearch.kubedb.com/es-demo-2 created

Verify Auto-backup configured

Now, let’s verify whether the auto-backup has been configured properly or not.

Verify Repository

At first, let’s verify whether Stash has created a Repository for our Elasticsearch or not.

❯ kubectl get repository -n demo-2
NAME            INTEGRITY   SIZE   SNAPSHOT-COUNT   LAST-SUCCESSFUL-BACKUP   AGE
app-es-demo-2                                                                25s

Now, let’s check the YAML of the Repository.

❯ kubectl get repository -n demo-2 app-es-demo-2 -o yaml
apiVersion: stash.appscode.com/v1alpha1
kind: Repository
metadata:
  name: app-es-demo-2
  namespace: demo-2
  ...
spec:
  backend:
    gcs:
      bucket: stash-testing
      prefix: stash-backup/demo-2/elasticsearch/es-demo-2
    storageSecretName: gcs-secret

Here, you can see that Stash has resolved the variables in prefix field and substituted them with the equivalent information from this new database.

Verify BackupConfiguration

Now, let’s verify whether Stash has created a BackupConfiguration for our Elasticsearch or not.

❯ kubectl get backupconfiguration -n demo-2
NAME            TASK                            SCHEDULE      PAUSED   AGE
app-es-demo-2   elasticsearch-backup-6.3.0-v7   */3 * * * *            77s

Now, let’s check the YAML of the BackupConfiguration.

❯ kubectl get backupconfiguration -n demo-2 app-es-demo-2 -o yaml
apiVersion: stash.appscode.com/v1beta1
kind: BackupConfiguration
metadata:
  name: app-es-demo-2
  namespace: demo-2
  ...
spec:
  driver: Restic
  interimVolumeTemplate:
    metadata:
      name: elasticsearch-es-demo-2
    spec:
      accessModes:
      - ReadWriteOnce
      resources:
        requests:
          storage: 1Gi
      storageClassName: standard
    status: {}
  repository:
    name: app-es-demo-2
  retentionPolicy:
    keepLast: 5
    name: keep-last-5
    prune: true
  runtimeSettings: {}
  schedule: '*/3 * * * *'
  target:
    ref:
      apiVersion: appcatalog.appscode.com/v1alpha1
      kind: AppBinding
      name: es-demo-2
  task:
    name: elasticsearch-backup-6.3.0-v7
  tempDir: {}
status:
  conditions:
  - lastTransitionTime: "2021-02-12T12:24:07Z"
    message: Repository demo-2/app-es-demo-2 exist.
    reason: RepositoryAvailable
    status: "True"
    type: RepositoryFound
  - lastTransitionTime: "2021-02-12T12:24:07Z"
    message: Backend Secret demo-2/gcs-secret exist.
    reason: BackendSecretAvailable
    status: "True"
    type: BackendSecretFound
  - lastTransitionTime: "2021-02-12T12:24:07Z"
    message: Backup target appcatalog.appscode.com/v1alpha1 appbinding/es-demo-2 found.
    reason: TargetAvailable
    status: "True"
    type: BackupTargetFound
  - lastTransitionTime: "2021-02-12T12:24:07Z"
    message: Successfully created backup triggering CronJob.
    reason: CronJobCreationSucceeded
    status: "True"
    type: CronJobCreated
  observedGeneration: 1

Notice the schedule section. This time the BackupConfiguration has been created with the schedule we have provided via annotation.

Also, notice the target section. Stash has automatically added the new Elasticsearch as the target of this BackupConfiguration.

Verify Backup

Now, let’s wait for a backup run to complete. You can watch for BackupSession as below,

❯ kubectl get backupsession -n demo-2 -w
NAME                       INVOKER-TYPE          INVOKER-NAME    PHASE       AGE
app-es-demo-2-1613132831   BackupConfiguration   app-es-demo-2               0s
app-es-demo-2-1613132831   BackupConfiguration   app-es-demo-2   Running     17s
app-es-demo-2-1613132831   BackupConfiguration   app-es-demo-2   Succeeded   41s

Once the backup has been completed successfully, you should see that Stash has created a new directory as pointed by the prefix field of the new Repository and stored the backed up data there.

Backup data in GCS Bucket
Fig: Backup data in GCS Bucket

Auto-backup with custom parameters

In this section, we are going to backup an Elasticsearch database of demo-3 namespace. This time, we are going to pass some parameters for the Task through the annotations.

Create Storage Secret

At first, let’s create the gcs-secret in demo-3 namespace with the access credentials to our GCS bucket.

❯ kubectl create secret generic -n demo-3 gcs-secret \
    --from-file=./RESTIC_PASSWORD \
    --from-file=./GOOGLE_PROJECT_ID \
    --from-file=./GOOGLE_SERVICE_ACCOUNT_JSON_KEY
secret/gcs-secret created

Create Database

Now, we are going to create an Elasticsearch CRO in demo-3 namespace. Below is the YAML of the Elasticsearch object that we are going to create,

apiVersion: kubedb.com/v1alpha2
kind: Elasticsearch
metadata:
  name: es-demo-3
  namespace: demo-3
  annotations:
    stash.appscode.com/backup-blueprint: elasticsearch-backup-template
    params.stash.appscode.com/args: --ignoreType=settings,template
spec:
  version: 7.9.1-xpack
  replicas: 1
  storageType: Durable
  storage:
    resources:
      requests:
        storage: 1Gi
    storageClassName: standard
  terminationPolicy: WipeOut

Notice the annotations section. This time, we have passed an argument via params.stash.appscode.com/args annotation along with the stash.appscode.com/backup-blueprint annotation.

Let’s create the above Elasticsearch CRO,

❯ kubectl apply -f https://github.com/stashed/elasticsearch/raw/6.3.0-v7/docs/auto-backup/examples/es-demo-3.yaml
elasticsearch.kubedb.com/es-demo-3 created

Verify Auto-backup configured

Now, let’s verify whether the auto-backup resources has been created or not.

Verify Repository

At first, let’s verify whether Stash has created a Repository for our Elasticsearch or not.

❯ kubectl get repository -n demo-3
NAME            INTEGRITY   SIZE   SNAPSHOT-COUNT   LAST-SUCCESSFUL-BACKUP   AGE
app-es-demo-3                                                                23s

Now, let’s check the YAML of the Repository.

❯ kubectl get repository -n demo-3 app-es-demo-3 -o yaml
apiVersion: stash.appscode.com/v1alpha1
kind: Repository
metadata:
  name: app-es-demo-3
  namespace: demo-3
  ...
spec:
  backend:
    gcs:
      bucket: stash-testing
      prefix: stash-backup/demo-3/elasticsearch/es-demo-3
    storageSecretName: gcs-secret

Here, you can see that Stash has resolved the variables in prefix field and substituted them with the equivalent information from this new database.

Verify BackupConfiguration

Now, let’s verify whether Stash has created a BackupConfiguration for our Elasticsearch or not.

❯ kubectl get backupconfiguration -n demo-3
NAME            TASK                            SCHEDULE      PAUSED   AGE
app-es-demo-3   elasticsearch-backup-6.3.0-v7   */5 * * * *            84s

Now, let’s check the YAML of the BackupConfiguration.

❯ kubectl get backupconfiguration -n demo-3 app-es-demo-3 -o yaml
apiVersion: stash.appscode.com/v1beta1
kind: BackupConfiguration
metadata:
  name: app-es-demo-3
  namespace: demo-3
  ...
spec:
  driver: Restic
  interimVolumeTemplate:
    metadata:
      name: elasticsearch-es-demo-3
    spec:
      accessModes:
      - ReadWriteOnce
      resources:
        requests:
          storage: 1Gi
      storageClassName: standard
    status: {}
  repository:
    name: app-es-demo-3
  retentionPolicy:
    keepLast: 5
    name: keep-last-5
    prune: true
  runtimeSettings: {}
  schedule: '*/5 * * * *'
  target:
    ref:
      apiVersion: appcatalog.appscode.com/v1alpha1
      kind: AppBinding
      name: es-demo-3
  task:
    name: elasticsearch-backup-6.3.0-v7
    params:
    - name: args
      value: --ignoreType=settings,template
  tempDir: {}
status:
  conditions:
  - lastTransitionTime: "2021-02-12T12:39:14Z"
    message: Repository demo-3/app-es-demo-3 exist.
    reason: RepositoryAvailable
    status: "True"
    type: RepositoryFound
  - lastTransitionTime: "2021-02-12T12:39:14Z"
    message: Backend Secret demo-3/gcs-secret exist.
    reason: BackendSecretAvailable
    status: "True"
    type: BackendSecretFound
  - lastTransitionTime: "2021-02-12T12:39:14Z"
    message: Backup target appcatalog.appscode.com/v1alpha1 appbinding/es-demo-3 found.
    reason: TargetAvailable
    status: "True"
    type: BackupTargetFound
  - lastTransitionTime: "2021-02-12T12:39:14Z"
    message: Successfully created backup triggering CronJob.
    reason: CronJobCreationSucceeded
    status: "True"
    type: CronJobCreated
  observedGeneration: 1

Notice the task section. The args parameter that we had passed via annotations has been added to the params section.

Also, notice the target section. Stash has automatically added the new Elasticsearch as the target of this BackupConfiguration.

Verify Backup

Now, let’s wait for a backup run to complete. You can watch for BackupSession as below,

❯ kubectl get backupsession -n demo-3 -w
NAME                       INVOKER-TYPE          INVOKER-NAME    PHASE       AGE
app-es-demo-3-1613133604   BackupConfiguration   app-es-demo-3               0s
app-es-demo-3-1613133604   BackupConfiguration   app-es-demo-3   Running     5s
app-es-demo-3-1613133604   BackupConfiguration   app-es-demo-3   Succeeded   48s

Once the backup has been completed successfully, you should see that Stash has created a new directory as pointed by the prefix field of the new Repository and stored the backed up data there.

Backup data in GCS Bucket
Fig: Backup data in GCS Bucket

Cleanup

To cleanup the resources crated by this tutorial, run the following commands,

❯ kubectl delete -f https://github.com/stashed/elasticsearch/raw/6.3.0-v7/docs/auto-backup/examples/
backupblueprint.stash.appscode.com "elasticsearch-backup-template" deleted
elasticsearch.kubedb.com "es-demo-2" deleted
elasticsearch.kubedb.com "es-demo-3" deleted
elasticsearch.kubedb.com "es-demo" deleted

❯ kubectl delete repository -n demo --all
repository.stash.appscode.com "app-es-demo" deleted
❯ kubectl delete repository -n demo-2 --all
repository.stash.appscode.com "app-es-demo-2" deleted
❯ kubectl delete repository -n demo-3 --all
repository.stash.appscode.com "app-es-demo-3" deleted