The system has a safe and reliable method of recovering lost information from your databases. The Repair process is activated when one of your databases is deactivated (whether it is because of hardware or a software problem or by the administrators).

During the time when your database is inactive, the requests which change the contents of the information (Insert/Update/Delete requests) are saved as “fail log” files. These are special binary files which contain the information for the unfinished requests. When the database is activated again in working mode, the system uses the information from the binary files to restore the request and to perform it in the recovered database. If more than one request has been processed during the time when the database was inactive, these are performed in the query they were registered as “fail log” files.

After the repair process is complete, the databases are synchronized again and the system can continue its normal working mode. There is additional security added if someone tries to repeat the execution of some of the “fail log” files (whether by mistake or on purpose, trying to destroy the synchronization of the databases).