We are currently using MySource Matrix v3.28.2, although I doubt that this is version dependent.
We have been noticing for some time that our backups have been increasing in size and are now becoming too large. We will hit the wall shortly on available disk space and we can't afford not to have regular backups. Adding further disk space is not an option.
What we are thinking is that the MySource Matrix file versioning system may be what is causing most of this backup bloat.
Once assets have been deleted from the asset map, they are for all intents and purposes, no longer available within MySource Matrix. However it appears that the file versioning system is keeping copies of the associated documents even though they are of no further use.
Given we have a lot of editors that appear to upload and delete files without a lot of thought (they may upload a file into the wrong location, delete it, upload again, delete it, repeat ad-nauseum, …), this results in a lot of file system wastage just to keep track of needless files. Trying to re-educate editors is a waste of time.
Can anyone shed any light on any downside to removing the unwanted files from the file_repository directory and perhaps cleaning up the sq_file_vers* tables? Is there a script that will do this for us?
Thanks in advance.
In the scripts directory there is a script: system_integrity_clean_old_files.php which deletes the old checked-out files from the data directory for file asset type.
The downside is the fact that some clients need complete history recorded for everything happened in the system for audit purpose.
If user can upload a file, and remove it without a trace, it would be hard to keep track of what happened.
But in your case, it might be ok to just remove those deleted file to save disk space. If you don't need to check roll-back history or keep records.
If you have rollback enabled, also take a look at: http://manuals.matrix.squizsuite.net/server-administrator/chapters/rollback-management-script/
If you don't need the rollback data, this script will help shrink the size of your database.
Thanks guys, however unless I am missing something, there are still some copies of "old" files in the data/file_repository directories and there are still references within the appropriate database tables.
Upon first examination, my assumption is that the "latest" copy of a deleted file is kept in the file system, along with the respective version information in the database.
I have run the system_integrity_clean_old_files.php script and the rollback_management.php script. Both scripts have removed some of the extraneous data.
Is there an approved way to remove the rest of the unrequired data from both the file system and the database?
[quote]
Is there an approved way to remove the rest of the unrequired data from both the file system and the database?
[/quote]
Do you include your cache directory in the backups by any chance?
Good thought, however the cache directory is not included in the backup.