Restoring Data

There may be a few scenarios when you need to restore data. In the event that:

The database is lost: The administrator must restore the last full backup of the database and apply the redo logs (Oracle) or apply incremental backups (Microsoft SQL Server) to roll forward the database to the vault time.
The vault is lost: It is very important to take a backup of the database in its current state, including the transaction logs (redo logs), before performing any restoring.
The vault (or both the vault and database) are lost: The server administrator must restore both the database and the vault from the last backup. After restoring the online database backup, the database has to be rolled forward to the vault backup time.

Note: If you do not have transaction logs (redo logs) available, this can cause data loss and limit your disaster recovery capabilities.

Reload the database from the last backup.
Simultaneously, reload the repository information (server configuration directory) from the last backup. If the full and incremental backups were used, then you can reload the last full backup, and in parallel, reload all subsequent incremental backups.
Restore the starteam-server-configs.xml file.
Start the server configuration.
Test the server configuration.

When all loads are complete, the repository is ready to use. It is okay if the archive or attachment folders have “future” files not represented in the database. Atlas Hub ignores the “future” files and, if those file revisions are eventually added again, Atlas Hub overwrites the existing files.