Backing Up and Restoring the Database Overview
As System Administrator, you can perform Junos Space Network Management Platform database backup, restore, and delete operations. Junos Space Network Management Platform enables you to back up the complete system data, which includes the MySQL database, the Cassandra database, and the network-monitoring database (containing the PostgreSQL data, configuration files, and performance data files). Because of this feature, if a system crashes, you can add a new system (Return Material Authorization (RMA)) and restore the configuration that existed in the crashed system from the backup file.
To perform database backup or restore operations, you must be assigned the System Administrator role. Only a System Administrator can initiate a backup operation from the Administration > Database Backup and Restore workspace.
When you initiate a backup operation, all databases are backed up by default. Because the network-monitoring database could be fairly large in size, you can select whether or not to back up this database from the Junos Space GUI by clearing the Network Monitoring check box from the Database Backup page (Administration > Database Backup and Restore > Database Backup). If sufficient disk space is unavailable, Junos Space Network Management Platform throws an error. Duration of the backup job might vary depending on the database size.
Junos Space Network Management Platform allows you to perform backup and restore operations even when the network-monitoring service is turned off.
If you have the Cassandra service running on at least one node in the fabric, the Cassandra database is backed up by default. If you do not want the Cassandra database to be backed up, you can clear the Cassandra check box from the Database Backup page (Administration > Database Backup and Restore > Database Backup).
In Junos Space Release 13.1 and earlier, a local backup operation
saves the backup file of the Junos Space database to a specific folder
/var/cache/jboss/backup) on the
active node. As an administrator, you may want the backup files to
exist on both the primary and secondary nodes so that when one of
the nodes crashes you can restore the system from the backup file
saved on the other node. In
this release, backup is initiated on the secondary node and the backup
file is saved to the default location (
/var/cache/jboss/backup) on the secondary node. If the backup operation is successful, then
the backup file is synchronized with (copied to) the primary node.
The following are the advantages:
The backup file is present on both the primary and secondary nodes due to which you can restore the system if one of the nodes crashes or is corrupted.
System performance of the primary node is not impacted because the backup operation is initiated on the secondary node.
When dedicated database nodes are present in the Junos Space fabric, database backup files are always stored in the dedicated database nodes. The database backups created before dedicated database nodes are added remain in the old nodes; the old backups are not moved to the dedicated database nodes. You can restore the system configuration from the old backup files even when later backups are present in the dedicated database nodes.
For disaster recovery, different additional database backup and restoration provisions must be made.
Restore the Junos Space Network Management Platform database if any of the following issues occur:
Junos Space Network Management Platform data is corrupted and you need to replace it with uncorrupted data.
The Junos Space Network Management Platform software is corrupted and you reinstalled the Junos Space Network Management Platform software.
You can restore a Junos Space database from a backup that is taken in the same release version only. For example, you can restore a Junos Space Release xx database only from a backup that is taken in Junos Space Release xx, where xx represents the version number.
In a multinode setup, the same backup file can exist on both the primary and secondary nodes. In such cases, when you choose to restore a system from a local backup file, Junos Space Network Management Platform randomly chooses a backup file from one of the nodes to restore the system.
Backing Up a Database
By default, Junos Space Network Management Platform automatically backs up the database once a week. However, the administrator can schedule a backup to run at anytime and perform either local or remote backup operations. All jobs that are completed before the start of the backup operation are captured in the database backup file.
During a backup operation, Junos Space Network Management Platform archives data files and the logical logs that record database transactions, such as the users, nodes, devices, and added or deleted services in Junos Space Network Management Platform.
The administrator can perform a local or remote database backup
operation. When the administrator performs a local backup operation,
Junos Space Network Management Platform backs up all database data
and log files to a local default directory
/var/cache/jboss/backup. You cannot specify a different database backup file location for
a local backup. No such restriction exists when backing up to a remote
For a remote backup, use only a Linux-based server. You must specify a remote host that is configured to run the Linux Secure Copy Protocol (SCP) command. You must also specify a valid user ID and password for the remote host. To ensure that you are using a valid directory, check the destination directory before you initiate a database backup operation to the remote system.
For instructions on how to back up the Junos Space Network Management Platform database, see Backing Up the Junos Space Network Management Platform Database.
Restoring a Database
When the System Administrator performs a restore database operation, data from a previous database backup is used to restore the Junos Space Network Management Platform database to its previous state. The administrator can restore the database through the Administration > Database Backup and Restore workspace (see Restoring the Junos Space Network Management Platform Database).
The restore database operation is performed while Junos Space Network Management Platform is in maintenance-mode. The system is therefore down on all nodes in the fabric and only the Web proxy is running. During this time, all Junos Space users, except the maintenance-mode administrator, are locked out of the Junos Space Network Management Platform.
After the Junos Space Network Management Platform database is restored, the Security Design database must be manually reindexed. For more information about Security Design, see the Security Design documentation.