Server Replication and Failover Procedures
Changing a Slave to a Master
To allow for data updates to continue even after a Master Server goes down, you can designate a Slave Server as your new Master Server. Once changed, the new Master Server will update any remaining Slave Servers just as the original Master did previously.
If you bring down a Slave to make it a Master, TT User Setup will be down temporarily as the change is being made, unless you have multiple Slave Servers running in your environment.
For disaster recovery purposes, or if there is an unrecoverable error on the Master Server, you can update your Slave Server to function as a Master by performing the following procedure.
To change a Slave Server to a Master Server
- On the Slave machine, run the TT User Setup Utility.
- Select the Edit Server Configuration File option (option 3).
- In the Database Replication Role section, select Master to change the database role.
- Click Save.
You may also choose to replace the contents of the Slave Server's TTUserSetupServer.ini file with the contents backed up from the Master Server's file.
Restoring Master Server Data from a Backup
You can restore data from a backup in the event that a Master Server goes down. All servers make backup copies of their local database just before starting so that such recovery is possible.
To ensure full system recovery of all your data, you must back up your TTUserSetupServer.ini file and the <database filename>.mdb file, which must be enabled in the TTUserSetupServer.ini file.
To restore data from a backup when a Master Server fails:
- If necessary, install a new TT User Setup Server.
- Select a location on the machine on which you've just installed the new server, and move the backed up <databaseName>.mdb file to that location.
- Copy the backup TTUserSetupServer.ini file to the <root drive>:tttt_user_setup directory on your machine.
- Double-click the TTUserSetupServer.ini file to open it.
- Ensure that paths, settings, and/or IP addresses are correct.
- Save the .ini file.
- Start your server.
When restoring from an old file, any changes that you made since that file was saved will be lost.
Compacting and Restoring a Database
Data corruption may result from a large increase in the size of a database or from repeated modifications to that database. As such, it is recommended that a database be regularly compacted.
To prevent such corruption, the TT User Setup Server's database is compacted automatically whenever the server is started. TT User Setup is designed so that TT Chron restarts the server weekly. However, the database may be compacted at any time by manually stopping and starting the server.
Only the original database is compacted when the server starts. Any backup databases are unaffected.
Take these additional steps to further help prevent database corruption:
- Never install a database on a shared drive.
- Open one connection to a database at a time only (e.g., Do not open Microsoft Access while the database is in use).
- Routinely compact the database (aside from server startup).
Should a database become corrupt, the following options are available for restoring or repairing the database:
- Restore from a recent backup database or a Slave's copy of the database. For detailed descriptions on Slave database replication and database restoration, refer to the following topics contained in this manual: Restoring Master Server Data from a Backup and Changing a Slave.
- Repair using the Microsoft Access compact utility or other available tools such as the Jet compact freely distributable utility http://support.microsoft.com/kb/295334