Should I migrate my data from a shared folder to a dataset

[path]

I just migrated to 11.2 and it looks like there are guides to port all familiar plugins over. I've noticed the canonical guides seem to suggest using a dataset for media.

Should I migrate all my data into a dataset? What's the best way to do this?

 

[Basil Hendroff]

I currently have all my media stored in folders within a media dataset. The UI reports how much space the media occupies, but I don't get any immediate feedback into how much space different types of media are occupying. So, yes, I think it's a good idea to place different types of media into their own dataset. Doing so, you will get more reporting information from the UI and a better insight into the types of media you own. It would still be a good idea to group these datasets together, by setting up them up as child datasets under a root. eg. media dataset with child datasets movies, music, pictures, video, etc. This way, the UI still informs you of the total space all media occupies.

The other advantage of separating different types media into their own datasets is that an application may only be concerned with one type of media eg. for something like Plex, all media datasets are mount point candidates, but for a music player plugin, only the music dataset would be applicable. You get this type of granularity using separate datasets for media type over a single media dataset.

Note: What's stopping me from doing what I suggest and not what I actually do is that I'm too far down the track to change things. For instance, if I did, Plex would refresh all its metadata and I would have have to go in and manually tidy things up. With a substantial library, this is not a trivial task.

 

[path]

Disk space usage monitoring is not very useful to me. And when I need it I can just use the command line. I understand the granularity advantage. Any other reasons? Anything to do with permissioning?

 

[Basil Hendroff]

Anything to do with permissioning?

Not that I can immediately see, but refer to the appended note in my original post. I've provided a reason why you might not want to change from current arrangements.

Disk space usage monitoring is not very useful to me. And when I need it I can just use the command line.

Slow at best. Impractical for large libraries. The UI provides immediate feedback.

 

[path]

Thank you! I'll stick with what I have.

 

[pro lamer]

With separate datasets one can:
1. Set different allocation unit size for each of them (default is 128K). In theory it can save some space but I think the gain would be minimal (edit: at least for media)

2. Can have separate snapshot sets

I guess some more but these are what I'm saying from the top of my head.

Sent from my phone