Our company was consolidating a dozen data centers to three new data centers. Our company retains several petabytes of data, in archive, for our customers and the old hardware running TSM 5.5 was not adequate for our needs. Rather than upgrade our existing servers and TSM version we built three new servers running TSM 7.1.100 and exported the data to three new instances.
Backups were being captured using a combination of TSM, Backup Exec, and Commvault. All backup data was being retained forever which was not sustainable. Working with our business leaders we developed backup policies based upon each division and performed a discovery to identify what needed to be backed up. New Infrastructure running TSM 7.1.100 was built out to accomodate our backup requirements as well.
The archive environment was on an extremely tight deadline. The data in "archive" was really just backed up and was not well organized. If I had to do it again I would have restored the data and archived it using the processes we have developed since then. The new process gathers all data for a particular project into a single archive. The description for the archive contains the project code, ticket number for the archive request, and the data of the archive. This makes identifying all data for a given project easy to identify and retrieve.