|Attention: DocWiki has reached EOL and will be decommissioned January 25, 2019.|
Cisco Unified MeetingPlace, Release 7.0 -- About Database Backups, Archives, and Restoration
Main page: Cisco Unified MeetingPlace, Release 7.0
Caution!! We recommend that you do not run reports or backups during peak hours of use, such as on the hour. Doing so may cause the system to fail.
We highly recommend that you back up and archive the Cisco Unified MeetingPlace Application Server for deployments with both a single Application Server and a deployment with Application Servers deployed for failover. You must set up the backup process on each Application Server individually.
The database backups include IP addresses and hostnames; therefore, you can only restore the backup onto the same Application Server where the backup occurred.
There are three types of database backups:
- L0 (Level 0) backup. This is the most common database backup. This is a complete physical and logical backup of the database from which data can be restored.
- L1 (Level 1) backup. The L1 backup is an incremental backup. It contains a backup of all the data that has been changed since the last L0 backup. It takes much less disk space than an L0 backup; however, it cannot be used for full restoration. If the system fails, you must use both the L0 and L1 backup files to restore data.
- L2 (Level 2) backup. The L2 backup is incremental to the L1 backup, so it needs both the L0 and the L1 backups to restore data.
Cisco Unified MeetingPlace uses a combination of L0, L1, and L2 backups and uses an Informix command called ontape for the backup mechanism.
The database backup file is physically located on the system disk of the Application Server. The system disk can contain up to three automatically-created L0 backups: the current L0, plus the previous one or two L0 backups. The L1 and L2 backups are also kept there. All of the older backups are removed from the system disk during the cleanup process.
Caution! Use caution if you manually modify the backup files on the local disk or in the archive location. For successful data restoration, the three levels of backup files must be present in the correct order. For example, if the correct L0 and L2 backup files are present while the appropriate L1 backup file is missing, then the data cannot be restored.
You can enable or disable an automatic backup. If the automatic backup is enabled, an L0 backup happens twice a week, every Monday and Thursday at 11:00PM, local server time. The L1 backup is run each day at 1:00AM, local server time, while the L2 backups are done daily at 4:00AM, 8:00AM, 12:00PM, 4:00PM, and 8:00PM, local server time. The schedule is stored in the crontab file.
Note: Advanced system administrators can change the frequency of the automatic backups by editing the crontab file. Be careful when modifying the cron schedule, which determines the order of the backups.
The automatic backup process also incorporates archiving (if enabled) and cleanup. This ensures that if there is a database corruption or disk failure, in the worst case, less than five hours of data is lost.
Cleanup Process for Database Backups
The cleanup process occurs before every scheduled backup in the crontab file. During the cleanup process, the following files are deleted:
- Backup files older than seven days.
- Unusable files, such as L1 and L2 backup files that are older than the oldest remaining L0 backup file.
Note: If you disable automatic backups, the cleanup process continues to run as scheduled in the crontab file. Therefore, if you want to keep backup files that are older than seven days, you must archive them.
About Archiving the Database Backup Files and Other Files
Archiving makes a remote copy of all the backup files and external files, such as licenses, meeting attachments, and voice recordings that have not yet been deleted by the automatic system cleanup processes. Backup files and archives do not include some configuration settings, SNMP configuration settings, or SMTP configuration settings. If a newly archived file has the same name as an existing archived file, the new file overwrites the old file. Maintaining the archive and the remote system used for storing the archive is the responsibility of the system administrator.
Note: The Cisco Unified MeetingPlace archiving process does not delete files from the remote archiving location. This is to prevent data loss and to permit the restoration of files from any point in time in which a backup exists. Since files are never deleted, it is possible for you to be unable to write to the remote file location due to insufficient disk space or because it has reached the maximum number of files in a single folder. We strongly recommend that you periodically delete content from the remote archive location to ensure that there is enough space for your needs.
Automatic archiving can be enabled or disabled. When enabled, it is initiated by and happens after the automatic database backup. There are two options:
SSH/rsync Archiving Method (Recommended)
The remote server to which you archive files must support rsync and SSH connections:
- To archive to a UNIX or Linux server, SSH service and rsync must be enabled on that server. Both SSH service and rsync are included in most UNIX and Linux software distributions.
- To archive to a Windows-based server, both an SSH server and an rsync utility must be installed on that server. There are many commercial third party SSH and rsynch utility software programs available. You must provide and install one on your Windows-based server prior to setting up Cisco Unified MeetingPlace archives.
FTP Archiving Method
The following restrictions apply to the FTP archiving method:
- The FTP archiving method does not use a secure connection to transfer files to the remote server.
- The FTP archiving method enables the Cisco Unified MeetingPlace Application Server to transfer backup files and other critical files to the remote Application Server; FTP clients cannot transfer files to the Cisco Unified MeetingPlace Application Server.
- Make sure that the remote host login credentials provide the permissions required to create new directories within in the directory specified in the Pathname location of archive field. For example, if you enter "pub" in the Pathname location of archive field, the following directories are automatically created when the archiving script runs:
The destination archive directory on the remote server may contain files or directories that are not present in the Cisco Unified MeetingPlace directory that is being archived. Those extra files and directories in the remote server will be left undisturbed.
Therefore, when archiving information to an FTP server, you may see messages such as the following:
- Old directory `test' is not removed
- Old directory `mpx-record/conf/001005' is not removed
- Old directory `mpx-record/conf/001192' is not removed
- Old file `tape_2009-06-19-16-00-02-Level-2.gz' is not removed
- Old file `tape_2009-06-19-20-00-02-Level-2.gz' is not removed
These messages are for information only and refer to the FTP server file system. They do not refer to the Cisco Unified MeetingPlace Application Server. It is the responsibility of the network administrator of the remote server to manage the archive directory on that server.