Table of Contents

Driven Administrator Guide

version 2.1.4

Backing Up and Restoring Driven Repositories

To be prepared for disaster recovery, you must implement a policy for taking regular snapshots of the Driven repository. Use either Amazon S3 or a filesystem repository for the backup-and-restore system.

Prerequisites for Backup to S3 Repos

If you want to back up your data to an Amazon S3 repository, the procedure for configuring Amazon Web Services (AWS) credentials depends on whether there is an embedded or an external Elasticsearch datastore for the Driven deployment.

Embedded Elasticsearch Datastore: Either configure the following parameters of the driven.properties file:

driven.aws.accesskey=
driven.aws.secretkey=

Or set the following environment variables:

AWS_ACCESS_KEY_ID
AWS_SECRET_KEY

External Elasticsearch Cluster: The AWS credentials are set in the Elasticsearch layer, not part of the Driven configuration. Install the Amazon Web Services Cloud plugin on the Elasticsearch cluster. The AWS credentials can be configured in several different ways, including setting properties in the YML file, environment variables, or IAM roles. See the AWS Cloud Plugin page on GitHub for more information.

The backup Command

Use the Driven CLI Client to administer backup and recovery functions. The following CLI example shows information about options that can be used with the backup command.

$ driven backup --help

java driven.management.backup.Backup [options...]

Optional:
 env vars: DRIVEN_CLUSTER, DRIVEN_HOSTS

Option                       Description
------                       -----------
--bucket                     S3 bucket for repository
--cluster                    Driven cluster name (default: driven)
--debug [Boolean]            enable debugging (default: false)
--delete                     name of snapshot to delete from repository
--delete-older-than          delete snapshots older than supplied date ( yyyy-MM-dd )
--display-width <Integer>    width of display (default: 80)
--help
--hosts                      Driven server host(s) (default: localhost)
--json [Options$JsonOpts]    output data as json (default: values)
--list                       list repositories, or if --repository is used list
                               snapshots in repository
--no-header
--path                       specify filesystem path for repository
--prefix                     S3 prefix for repository bucket (default: snapshots)
--print                      print query parameters
--region                     AWS region for repository bucket (default: us-east-1)
--register-fs                register a local repository with name
--register-s3                register a remote S3 repository with name
--repository                 name of repository
--restore                    name of snapshot to restore from repository
--snapshot                   name of snapshot to create in repository
--unregister                 name of repository to unregister
--verbose                    logging level (default: info)

Backing Up Your Driven Data

The backup process generally entails creating a repository location and storing data snapshots there.

There are two ways to create a repository. You can either specify the repository in the driven.properties file, or you can use the Driven CLI Client to create the repository.

Step 1: - Option 1 Configure the driven.properties file to create a repository

To create only one default repository, add the following to the driven.properties file. Use either Amazon S3 or a filesystem repository:

For Amazon S3:

driven.backup.repository.type=s3
driven.backup.repository.bucket=myExampleBucketName
driven.backup.repository.region=us-east-example

For shared filesystem repository:

driven.backup.repository.type=fs
driven.backup.repository.path=/opt/driven-backup/snapshots

Step 1: - Option 2 Create a repository with the CLI Client

For Amazon S3 (servers in a cluster must have appropriate AWS credentials for the bucket):

$ driven backup --register-s3 s3-repo --bucket driven-backup

registered S3 repository 's3-repo' at '[us-east-example] driven-backup/snapshots'

For shared filesystem repository (all the servers in a cluster must have read/write access to the registered path):

$ driven backup --register-fs fs-repo --path /opt/driven-backup/snapshots

registered FS repository 'fs-repo' at:
     '/opt/driven-backup/snapshots'

To list the repositories that you created:

$ driven backup --list

Repository            Type Location
----------            -------------
s3-repo       	      s3   s3://driven-backup/snapshots
fs-repo               fs   /opt/driven-backup/snapshots

Step 2: Create a snapshot

$ driven backup --repository s3-repo --snapshot snapshot_3
CREATING.........................
Snapshot                  State  	  Status    Started                    	     Finished
--------                  -----  	  ------    -------                    	     --------
snapshot_3                SUCCESS	  OK        Thu Jun 19 10:35:08 PDT 2014     Thu Jun 19 10:35:33 PDT 2014

If no argument is given, a snapshot name with a universally unique identifier (UUID) is automatically assigned.

To list the snapshots that you created:

$ driven backup --repository s3-repo --list
Snapshot                 State  	Status  Started                    	   Finished
--------                 -----  	------  -------            	           --------
snapshot_1               SUCCESS	OK      Tue Jun 17 10:32:56 PDT 2014   Tue Jun 17 10:33:08 PDT 2014
snapshot_2               SUCCESS	OK      Tue Jun 17 11:33:02 PDT 2014   Tue Jun 17 11:33:15 PDT 2014

Restoring Data from a Snapshot

$ driven backup --repository s3-repo --restore snapshot_2
RESTORING........
snapshot snapshot_2 restore status 200

In the example above, "snapshot_2" is restored.