This topic describes how to backup and restore Redis Enterprise for Pivotal Platform (formerly Pivotal Cloud Foundry - PCF).

You must use the BOSH Backup and Restore (BBR) command-line tool to backup and restore your Pivotal Platform deployment. These backup and restore instructions apply specifically to your Redis Enterprise for Pivotal Platform deployment.

Installing the BBR command-line tool

To install the BBR command-line tool:

  1. Find the latest version of the BBR tool at:

    https://github.com/cloudfoundry-incubator/bosh-backup-and-restore/releases

  2. Right-click on the latest tar archive of the tool and copy the link address.

  3. To download the BBR tool, in the Ops Manager run:

    wget <archive_link_address> -O bbr.tar && tar -xvf bbr.tar

The bbr tool is extracted to /releases/bbr.

Backup Redis Enterprise for Pivotal Platform

Pivotal recommends that you run a daily backup.

Before you configure a backup process you must have:

  • BBR user credentials - Find the values for bbr_user and bbr_user_pass in the Ops Manager file:

    https://<host>/api/v0/deployed/director/credentials/uaa_bbr_client_credentials

  • Deployment name - On the Ops Manager, find the deployment name that begins with redis-enterprise- in the output of the command:

    bosh -e cf deployments

To backup your Redis Enterprise for Pivotal Platform deployment:

  1. Run the backup command:

    BOSH_CLIENT_SECRET=<bbr_user_pass> <path_to_bbr> releases/bbr deployment --debug --target <Bosh_OpsManager_IPaddress> --ca-cert=/var/tempest/workspaces/default/root_ca_certificate --username <bbr_user> --deployment <deployment_name> backup

    For example:

    root@ip-10-0-0-241:~# BOSH_CLIENT_SECRET=Nu5iIIXcXN-RIKVBk6JNoCOJVLN1kB6N releases/bbr deployment --debug --target 10.0.16.5 --ca-cert=/var/tempest/workspaces/default/root_ca_certificate --username bbr_client --deployment redis-enterprise-b25242b578838c375ba1 backup

    You can remove the –debug flag to reduce the command output.

    The backup is saved in the directory named with <deployment_name>_<date>, for example: redis-enterprise-b25242b578838c375ba1_20180312T135418Z

  2. Copy this backup archive from the Ops Manager machine to a dedicated storage location.

Restore Redis Enterprise for Pivotal Platform

To restore a Redis Enterprise for Pivotal Platform deployment:

  1. On the new Pivotal Platform Foundation, install a redis-enterprise tile with the same configuration of the original tile.

    Note:
    A different cluster name or number of machines than the original tile causes errors in the restore.

  2. In the cluster configuration tab, check the Recovery Mode checkbox and click Apply change.

  3. Copy the backup archive to the Ops Manager.

  4. Run the restore command to recover the cluster:

    Note:
    This command does not recover databases.
    BOSH_CLIENT_SECRET=<bbr_user_pass> <path_to_bbr> deployment --debug --target <Bosh_OpsManager_IPaddress> --ca-cert=/var/tempest/workspaces/default/root_ca_certificate --username <bbr_user> --deployment <deployment_name> restore --artifact-path=<archive_path>

    For example:

    BOSH_CLIENT_SECRET=sSvt2-_ykNHOX_0BW1_LImYlJdVEMDew bbr deployment --debug --target 10.0.16.5 --ca-cert=/var/tempest/workspaces/default/root_ca_certificate --username bbr_client --deployment redis-enterprise-4614f7015e079299ba4c restore --artifact-path=redis-enterprise-0b094c96bab6f58bb133_20180326T125355Z

  5. Connect to one of the cluster machines with SSH.

  6. To see that all of the nodes are connected, run: rladmin status extra all

  7. On each node, set the cluster recovery source path to the cluster backup file:

    rladmin node <node_id> recovery_path set <import_path>

    For example:

    rladmin node 1 recovery_path set /var/vcap/store/redis/redis/

    rladmin node 2 recovery_path set /var/vcap/store/redis/redis/

    rladmin node 3 recovery_path set /var/vcap/store/redis/redis/

  8. To verify that database recovery is ready, run:

    rladmin recover list

    Make sure there are no errors in the command output.

  9. To start database recovery, run:

    rladmin recover all

  10. In the Ops Manager, uncheck the Recovery Mode checkbox and click Apply changes.