You may need to understand some of the components in order to decide on this
1) Catalogs keep a record of exactly what has been backed up and where it was backed up to. The catalogs provide the restore selections and can be truncated (which results in losing the abuility to drill down to individual records but can still provide a DR restore at a volume or database level - AND a full catalog can be recreated if a truncation has occurred). If a piece of media is overwritten or reclaimed (deleted) then the catalogs for the media will be automatically deleted. Catalogs do use disk space (exact location can be customized). As catalogs will eventually be deleted by media rotation causing overwrites (for tapes) and DLM activity (for disk etc) then as long as the storage is large enough they should be self policing.
2) Job History info - this info is only in the database and is a summary of activity about when jobs ran. If a combination of job histories and job alerts gets too large then the Backup Exec Database can end up being larger than the limitations of SQL Express and cause problems. A lot of the reports within Backup Exec will use Job History information from inside the database.
3) Job Log info - this is more detailed information about the specifics of the job activity (and can also have the level of detail adjusted) and is stored in XML files in the data folder (as a defualt location)- which of course does mean disk space can be consumed if kept for too long.
The difference between the information in 2 and 3 can be seen by double-clicking a recent job (in the lower pane of the job monitor) and then comparing the two tabs.
Obviously if you have some kind of official audit process for backups then deleting Job logs and / or job Histories too soon might affect you (especially if you also run reports)
Another problem that we have sometimes seen is with the removal or truncation of one of more of the above is that if media (tapes) are sent off-site and the information about when the media was used and what was backed up to it is missing, then how would an admin work out what media is needed for a restore at a later date. Admittedly if you use a professional data storage company for your off-site storage, then most of them do provide reporting on dates media were sent to them which can help with this sort of problem.
We have also seen issues where a restore does not work where having the original backup job logs and histories might help us identify that something was not spotted at the time of the backup (or even that the backup job used more media than is available for the restore )
This does mean your question becomes a balancing act between what you might need to refer to in the future and resources on the BE server itself. and we don't really have a best practice. You might find that using the ability for Backup Exec to generate reports and e-mail them, could be used to at least maintain records outside of Backup Exec, that would mean you don't need to keep laborious manual records.
Looking in the very latest version of Backup Exec we have an option to "Keep Job History for data on media that have current overwrite protection periods" which would help with most of the above points and would link the job history very closely to when the catalogs get removed, and still leave you with needing to make a decision about Job Logs.
If it was me (so not an official best practice, just a suggestion) I would probably choose to do something like this:
- Let Catalogs delete themselves as and when media is overwritten and reclaimed (and I would put the catalogs on their own volume with plenty of space)
- Keep Job History Data against current overwrite periods (using the option in the later versions of BE)
- Keep Job Logs for the same length of time as the overwrite protection for your monthly backups. (which is likely to be longer than our default)
