Astra Connector (deprecated)

Requirements

Update Trident, activate ACP, install Astra Connector

The prerequisite is a current Trident (version 24.02). Depending on which version is currently on the cluster, this may need to be updated (e.g. if deployed via Helm: helm upgrade trident -n trident netapp-trident/trident-operator).

Then the ACP (Astra Control Provisioner) mode of Trident is activated, this enables the integration between Astra Control and Trident. Checkout the ACP documentation to install ACP!

Astra Connector must be installed in the Kubernetes cluster. I will describe how exactly in a new post. Until then, have a look at the official documentation .

Applications

Snapshots & Backup

First, you need the object store on which backups and metadata (also for snapshots) are stored. Here in the example for an Ontap-S3, for StorageGrid simply change the providerType to “storagegrid-s3”. In addition, change the “name” if necessary (if several buckets are used, you can differentiate between them using this field). Also adjust the “endpoint” and “bucketName”:

apiVersion: v1
kind: Secret
metadata:
  name: s3-creds
  namespace: astra-connector
type: Opaque
stringData:
  accessKeyID: <S3 access key>
  secretAccessKey: <S3 secret key>
---
apiVersion: astra.netapp.io/v1
kind: AppVault
metadata:
  name: my-appvault
  namespace: astra-connector
spec:
  providerType: ontap-s3
  providerConfig:
    endpoint: s3.company.org
    bucketName: astra
    skipCertValidation: "true"
  providerCredentials:
    accessKeyID:
      valueFromSecret:
        name: s3-creds
        key: accessKeyID
    secretAccessKey:
      valueFromSecret:
        name: s3-creds
        key: secretAccessKey

Now you can define applications (i.e. what should be saved together), here the “standard case”, an app corresponds to a namespace. Simply adjust “name” and “namespace”:

apiVersion: astra.netapp.io/v1
kind: Application
metadata:
  name: wordpress
  namespace: astra-connector
spec:
  includedNamespaces:
    - namespace: wordpress

Now you can take a snapshot. To do this, adjust “applicationRef” to the app name of the YAML above and, if necessary, the appVaultRef (if you have named it differently). Then check the status with kubectl get snapshot -n astra-connector

apiVersion: astra.netapp.io/v1
kind: Snapshot
metadata:
  name: snap1
  namespace: astra-connector
spec:
  applicationRef: wordpress
  appVaultRef: my-appvault

Very similar for a backup:

apiVersion: astra.netapp.io/v1
kind: Backup
metadata:
  name: bkp1
  namespace: astra-connector
spec:
  applicationRef: wordpress
  appVaultRef: my-appvault

And the best way to do this is of course to set it up as a schedule:

apiVersion: astra.netapp.io/v1
kind: Schedule
metadata:
  name: sched
  namespace: astra-connector
spec:
  applicationRef: wordpress
  appVaultRef: my-appvault
  backupRetention: "2"
  snapshotRetention: "1"
  granularity: hourly
  minute: "10"

Restore Application

Backup without restore would be kind of stupid… To do this, determine the backup path and use:

kubectl -n astra-connector get backup bkp1 -o=jsonpath='{.status.appArchivePath}'

Now you can restore your application!

apiVersion: astra.netapp.io/v1
kind: BackupInplaceRestore
metadata:
  name: bkprestore-to-wordpress
  namespace: astra-connector
spec:
  appVaultRef: my-appvault
  appArchivePath: <path>