The backup scripts must be tested to ensure that there are no unseen obstacles such as firewall rules or ARP caches on switches which could affect database connectivity or client connectivity to the server, so as to avoid unnecessary downtime and data loss. It is better to fail in a controlled test than to discover that the backup jobs did not run.
The formal backup/restore procedure must be documented and documentation must be available to the relevant staff. Staff must receive training and must practice the procedure at regular intervals. This will ensure that the procedure works as expected, and in the event of an outage of the master j5 server it is quite easy for staff to restore the service from the standby server.
When starting the j5 server on the production or standby machines, certain scheduled tasks will be run within j5 (for example, emailing of shift reports, etc). In order to prevent these from being run on a test-run of failover, it is possible to start the j5 services in SLAVE mode rather than MASTER mode. To do this, alter the local configuration file for the backup server to contain the line _mode =’SLAVE’. This should only be done for a test run, and not for the actual failover procedure.