Contents

1   Introduction

This document describes how to restore the configuration of Atlas for Cloud Execution Environment (CEE).

A restore of CEE Atlas is done by using the atlas command, executed from the Atlas Virtual Machine (VM). The input to restore the Atlas configuration is a previously taken backup, containing key configuration files.

For more information about the atlas command used for the restore procedure, refer to Atlas Software Management Guide.

Refer to the document Backup and Restore Overview for more information on Atlas restore, including the following:

• Backup contents
• Backup sizes and locations
• Retention policies
• Procedure durations
• Recommended backup strategy

For an overview of all backup and restore options available in CEE, refer to the Backup and Restore Overview.

1.1   Prerequisites

This section describes the prerequisites which must be fulfilled before Atlas can be restored.

1.1.1   Conditions

Before the restore procedure can be performed, the following conditions must apply:

• The user needs root access to do a restore, using the atlas command.
• A valid backup file must be available in a predefined persistent storage location.

2   Downloading Backup Files from Swift

This section describes how backups are downloaded, that is copied, from the AtlasBackups container in Swift.

To download a backup from Swift, do the following:

1. Log on to the Atlas VM.
2. Switch user from atlasadm to root:

   atlasadm@atlas:~$ sudo -i

3. Ensure that the current directory is set to the backup directory:

   root@atlas:~# cd /var/archives/

4. Perform the following steps to verify that the backup files to download are available in Swift:
   1. If the Atlas backup ID is known, continue with Step_c.
   2. If the Atlas backup ID is unknown, search the Atlas backup ID from the list of the available backups in Swift:

      root@atlas:/var/archives# swift list AtlasBackups -d / AtlasBackup1520490601/ AtlasBackup1520490602/

      The command output displays the Atlas backup directories in the following format: <atlas_backup_name><backup_id>

   3. List the files stored for the given backup ID:

      root@atlas:/var/archives/# swift list AtlasBackups | grep <backup_id>

      Where backup_id is the ID of the backup.

      See below an example of the command and the output:

      root@atlas:/var/archives/# swift list AtlasBackups | grep 1520490601

      AtlasBackup1520490601/.meta AtlasBackup1520490601/AtlasBackup.1520490601-all-mysql-databases.sql.bz2.enc AtlasBackup1520490601/AtlasBackup.1520490601-etc-puppet-hieradata-passwords.yaml.master.tar.gz.enc AtlasBackup1520490601/AtlasBackup.1520490601-home-atlasadm.master.tar.gz.enc AtlasBackup1520490601/AtlasBackup.1520490601-root.master.tar.gz.enc AtlasBackup1520490601/AtlasBackup.1520490601.sha256.enc a

5. Download the backup for a specified ID:

   root@atlas:/var/archives/# swift download AtlasBackups -p <atlas_backup_name><backup_id>

   root@atlas:/var/archives/# swift download ⇒
   AtlasBackups -p <atlas_backup_name><backup_id>

   where atlas_backup_name is the name of the backup and backup_id is the ID of the backup.

   And example of the command is the following:

   root@atlas:/var/archives# swift download AtlasBackups -p AtlasBackup1520490601

   root@atlas:/var/archives# swift download ⇒
   AtlasBackups -p AtlasBackup1520490601

3   Atlas Restore

The restore feature replaces key configuration files and folders with the content in the backup.

To restore Atlas, do the following:

1. Determine if a local backup file is available, using the following atlas command:

   atlasadm@atlas:~$ sudo atlas backup-list

2. Select the relevant one of the following alternatives:
   1. If a local backup file is available, continue with Step 3.
   2. If no local backup is available, do the following:

      Copy the desired backup from the predefined persistent storage location to the directory /var/archives.

      For information on how to copy the backup using Swift, see Section 2.

3. Do the following:

   atlasadm@atlas:~$ sudo atlas backup-restore --d <backup_id> --p <backup_password> atlasadm@atlas:~$ sudo atlas backup-restore --d <backup_id> ⇒ --p <backup_password>

   Example:

   atlasadm@atlas:~$ sudo atlas backup-restore --d 1520490601 --p <backup_password> atlasadm@atlas:~$ sudo atlas backup-restore --d 1520490601 ⇒ --p <backup_password>

   The following output is received:

   Tue 02 May 2017 11:37:48 INFO: Starting Atlas restore... Tue 02 May 2017 11:37:48 INFO: Restoring database... Tue 02 May 2017 11:37:49 INFO: Done. Tue 02 May 2017 11:37:49 INFO: Restoring etc-puppet-hieradata-passwords.yaml... Tue 02 May 2017 11:37:49 INFO: Done. Tue 02 May 2017 11:37:49 INFO: Restoring home-atlasadm... Tue 02 May 2017 11:37:49 INFO: Done. Tue 02 May 2017 11:37:49 INFO: Restoring root... Tue 02 May 2017 11:37:56 INFO: Done. Tue 02 May 2017 11:37:56 INFO: Verifying users... Tue 02 May 2017 11:38:29 INFO: Applying configuration... Tue 02 May 2017 11:38:47 INFO: Restarting services... Tue 02 May 2017 11:38:50 INFO: Done. Tue 02 May 2017 11:38:50 INFO: Successfully restored from backup:1520490601

4. Verify that Atlas is operational.

   Note:  

   For an example of one way to verify that Atlas is operational, refer to section Post-Upgrade Activities in Atlas SW Upgrade.