Hi, when deleting items from history, what is the difference between deleting “collection” versus “collection and elements”?
Welcome @shlab
Good question!
A collection is a folder (directory) that contains elements (each is a reference to a dataset).
Deleting a collection removes the folder.
Deleting an element removes the reference to a dataset
If a dataset no longer has any references in any history, then the file itself is removed.
Copies of the same exact dataset referenced across histories and collection are only still just one file.
Example: if a job failed, it is probably safe to remove the collection and the elements inside of it, because you do not care about about any of those results, and they are potentially taking up quota space. Same for data you just don’t care about anymore for other reasons.
Tip: if you are cleaning up results to sort of start over, I think it is much easier for a few reasons to use the Copy Datasets function to move the files you want to continue working with into a new history. Then, delete the original history but maybe don’t purge it yet. Wait a week or so to make sure that you still have everything you need. Then go in and purge that history when you are actually out of working space.
But you can also can also delete files as you go and use the Storage Dashboard to clean those up periodically. Just be aware that it is easy to accumulate a lot of purged datasets in a history and that can make it confusing. As long as you are not working for several years always in the same single unnamed history, you are probably fine.
Does this help? Let us know! 
1 Like