CML - Recovery procedure for master node
This section describes how to recover master node (1st installed node) of the CML cluster. The same procedure can be used for the installation with the single CML node - the only difference is that both databases shall be taken from the regular database backup.
If the master node is not operational for a longer period of time, you can use the procedure from CML - Printer settings in case of one CML node failure to sustain the printing functionality.
The procedure does not require shutdown of the running CML nodes in the cluster.
Warning
If you are going to recover the YSoft SafeQ on a new PC (meaning on a new installation of operating system or on a different server), you must request the license reactivation on a new hardware. For more information see chapter Transferring license to new system (hardware or operating system change).
This article consists of:
Requirements
Kindly prepare all the required data before starting with the master node (1st installed node) recovery:
SQDB5 database backup from the regular system backup
Cluster environment: skip this point as the database backup will be obtained from another member of the CML cluster later in the procedure.
SQDB5_SQDW database backup from the the master node (1st installed node)
Cluster environment: The backup must be 7 days old at the most. If the backup is older than 7 days, the statistical data between the last backup date and 7 days back from the current moment will be missing.
Configuration files backup from the 1st installed node, eventually master CML GUID
TEMP and SPOOL folder backup from the 1st installed node if they are available
required only in case that cluster-readSharedFolderJob is disabled
required only in case that you wish to restore the print jobs that were stored on this server
Folder containing scripts for scanning (in case that scan to script is used)
Step 1 - restore hardware after server failure (optional)
Restore hardware required for YSoft SafeQ server.
Step 2 - restore software after server failure
1 | Install one of the supported operating systems for YSoft SafeQ server
|
2 | Install one of the supported databases (or select the embedded database during installation of YSoft SafeQ in the next step)
|
Step 3 - prepare the environment
1 | Perform this step on all running CML servers
Verify synchronization Check that all data are synchronized between all running YSoft SafeQ nodes.
|
2 | Perform this step on all CML servers that remained functional (did not crash)
Using the program pgAdmin III or MS SQL management studio connect to the SQDB5 database:
2a. Find out the master node db_flag using the following query:
2b. Run the following queries. Replace x by the proper db_flag for the master node (obtained in the previous step) before running them
:
|
3 | Perform the database backup of SQDB5 database on CML2 (the second node of the cluster).
|
Step 4 - install and configure the new server
1 | Perform following step on the node which you are recovering: 1a. Perform the installation Install YSoft SafeQ in the new location according to Installing YSoft SafeQ CML instructions. Installation requirements:
1b. Restore databases Restore databases according to CML - Database restore instructions with following exceptions:
1c. Run these queries at master node on SQDB5
1d. Modify the database Using the program pgAdmin III or MS SQL management studio connect to the SQDB5 database:
1e. Using pgAdmin III or MS SQL management studio run following query:
1f. Replace the configuration files Overwrite existing configuration files with original configuration backup:
1g. Restore spooled jobs Restore your backup of temp and spool directory
1h. Start services on Master node: Open Services (e.g. Start => Run => services.msc) and start all YSoft SafeQ services in any order:
|
Step 5 - verification of the functionality
1 | Log in to the web interface (YSoft SafeQ Web interface => System => System information) of the master node.
|
2 | Test authentication/print/copy/scan. All functions must be working. Also accounting must return the correct data.
|
3 | In case of non-clustered environment (1CML only), cache on every ORS has to be deleted. For specific instructions see article ORS - How to delete corrupted cache.
|


