Please use the form below to provide your feedback. Because your feedback is valuable to us, the information you submit in this form is recorded in our issue tracking system (JIRA), which is publicly available. You can track the status of your feedback using the ticket number displayed in the dialog once you submit the form.

Restore a Cluster Backup

  • Capella Operational
  • how-to
    +
    You can restore a cluster backup in a disaster recovery situation to restore your cluster to a previous point in time.
    If your cluster uses Couchbase Server version 7.6.6 or later and is deployed on AWS or Azure, you can choose to replicate your backups to 2 regions, other than the region where your cluster is deployed. You can restore a replicated backup like any other backup for your cluster. Cluster backups for GCP clusters cannot be replicated to additional regions.

    Prerequisites

    • You have the Project Owner role for the project with the cluster you want to restore. If you have the Organization Owner role, you have Project Owner permissions.

      For more information about roles in Capella, see Manage Organizations and Access.

    • You have previously taken or scheduled a cluster backup.

    • If you use Customer Managed Encryption Keys (CMEK) on your cluster:

      • The Capella CMEK ID must still be available in your organization.

      • The Key ID for the backup you want to restore must be available and enabled in your Key Management System (KMS).

      To view the CMEK ID and Key IDs used by your backups, go to Backup  Cluster Backups and expand the entry for a backup.

    Procedure

    You can restore a cluster backup from:

    Restoring a cluster backup also deletes all cluster access credentials and allowed IP addresses on your cluster.

    Before you restore a cluster backup, use version 4 of the Management API to get a list of all available cluster access credentials and get a list of all allowed IP addresses on your cluster.

    You can use the list to recreate your cluster access credentials and allowed IP addresses later.

    If the number of nodes deployed in your cluster changed between your backup and restore operations, Capella reverts your cluster configuration to the number of nodes present in the backup. Take care when restoring backups to make sure you do not lose necessary node configuration changes.
    To restore a cluster backup from the Capella Management API, see Restore Backup.

    Restore a Cluster Backup From the Cluster Backup Tab

    If you restore from your cluster’s Backup tab, you can restore your backup to the same cluster.

    To restore a cluster backup to the same cluster:

    1. On the Operational Clusters page, select the cluster where you want to restore a backup.

    2. Go to Backup  Cluster Backups.

    3. In the list of cluster backups, find the backup you want to restore.

    4. Click Restore.

    5. Confirm that you want to restore the backup.

    6. Click Restore.

    Restore a Cluster Backup From the Project Backups Tab

    If you restore from your project’s Backups tab, you can restore your backup to the same or a different cluster in your project.

    To restore a backup to the same or a different cluster:

    1. Go to the Projects tab.

    2. Click the project that has the backup you want to restore.

    3. Go to the Backups tab.

    4. For the cluster that has the backup you want to restore, click See All Versions.

    5. For the specific backup you want to restore, click Restore.

    6. (Cross-Region Backups Only) If your cluster backup was a cross-region backup, in the Restore from list, select the specific region that you want to use for your restore.

    7. Do 1 of the following:

      1. To restore your backup to a new cluster, click Restore To New Cluster and enter a cluster name, description, CIDR block, and choose your Availability Zone configuration.

        For more information about cluster configuration settings, see Configure Your Cluster.

        When restoring a backup that has cross-region copies, if you choose a different region in Restore from than the region where you want to deploy a new cluster, Capella tries to restore a copy from the same region as the new cluster, first.

      2. To restore to an existing cluster, in the Restore To list, choose the cluster you want to restore to.

    8. Click Restore.

    9. Confirm that you want to restore the backup.

    10. Click Submit.

    You can only restore a backup to a different compatible cluster if both clusters are currently running Couchbase Server version 7.6.6 or later. If a cluster is running an earlier version of Couchbase Server, you cannot restore its backups to another cluster.

    For example, if you have 2 clusters, with the first cluster on Couchbase Server version 7.6.5 and the second on Couchbase Server version 7.6.6, you cannot restore a backup from the 7.6.5 cluster to the 7.6.6 cluster until you upgrade the 7.6.5 cluster.

    If you do upgrade a 7.6.5 cluster to a later version of Couchbase Server, backups taken while the cluster was on version 7.6.5 can still be restored to that same cluster. Restoring a backup taken on version 7.6.5 will temporarily restore the cluster to Couchbase Server version 7.6.5. You can run an upgrade to restore the cluster back to a later version at any point after the restore.

    Next Steps

    Capella starts the restore process for your cluster backup on your cluster. Your cluster will be unavailable until the restore completes, and this can take some time.

    All existing data on your cluster will be lost. Your cluster configuration may change to match the configuration of the cluster stored in the backup.

    The time it takes to restore your cluster is not considered to be downtime based on the Capella Cloud Service Availability agreement.

    You can view the details and status for all your cluster backup restores on the Cluster Backup Restores page.

    The Size of the cluster used size at the time of your backup is only an estimate. Capella calculates the cluster size information for a backup at the start of the backup process, based on the current size of your cluster. The actual size of a restored cluster may be different from this initial estimate.