First, don't be hating!  I think this can be a useful feature, and hopefully wouldn't be too difficult to implement.  It's easy enough to do manually, but User-Shares are all about making things easier, right?

 

Ok, so I might have some directories in which the contents are very important.  I'm thinking Home Movies, Pictures, etc.  I'd like to have an option where I can setup a User-Share, and select multiple drives to copy to.  Sounds like how it already works, right?  Well I want it to write the data to EVERY drive, not just one of them.  For example, I have a share called Pictures.  I have it setup to write to disk1, disk2, and disk3.  I write a file called SonPark1.jpg to the Pictures share.  The file now exists as /mnt/disk1/Pictures/SonPark1.jpg, /mnt/disk2/Pictures/SonPark1.jpg, and /mnt/disk3/Pictures/SonPark1.jpg.

 

Why would this be useful?

 

This further protects your data in case of multiple drive failure.  If I have four drives with one being parity, ALL THREE DATA DRIVES have to die at the same time before I'd lose any of my very important data.  It's also extremely easy to use, just copy important data to the special User-Share, the rest is done for you.

 

Why do we NOT need this?

 

Sure, there's a work-around.  The average user can simply copy files over three times, or more if they want to copy to more drives.  A linux guru can simply copy to one drive and then have rsync run on the server once a day or so to copy the files to all the other drives.

First, don't be hating!  I think this can be a useful feature, and hopefully wouldn't be too difficult to implement.  It's easy enough to do manually, but User-Shares are all about making things easier, right?

 

Ok, so I might have some directories in which the contents are very important.  I'm thinking Home Movies, Pictures, etc.  I'd like to have an option where I can setup a User-Share, and select multiple drives to copy to.  Sounds like how it already works, right?  Well I want it to write the data to EVERY drive, not just one of them.  For example, I have a share called Pictures.  I have it setup to write to disk1, disk2, and disk3.  I write a file called SonPark1.jpg to the Pictures share.  The file now exists as /mnt/disk1/Pictures/SonPark1.jpg, /mnt/disk2/Pictures/SonPark1.jpg, and /mnt/disk3/Pictures/SonPark1.jpg.

 

Why would this be useful?

 

This further protects your data in case of multiple drive failure.  If I have four drives with one being parity, ALL THREE DATA DRIVES have to die at the same time before I'd lose any of my very important data.  It's also extremely easy to use, just copy important data to the special User-Share, the rest is done for you.

 

Why do we NOT need this?

 

Sure, there's a work-around.  The average user can simply copy files over three times, or more if they want to copy to more drives.  A linux guru can simply copy to one drive and then have rsync run on the server once a day or so to copy the files to all the other drives.

Need to be careful with how the delete function is implemented here.  A single delete could delete all 3/4 copies of critical files.

 

Ok, sort of like software mirror or syncing but at higher level.

 

I remember there being software to do this under windows.

Everytime you updated a file in My Documents, it would synchronize that change somewhere else.

 

I think the feature has merit, but also it can be handled at the user level with rsync and a cron entry.

 

I think it would be more useful if we had an easy way to remove a drive from the array without a full parity resync.

Then one of those "spare sync'ed" drives could be removed and sent offsite.

 

 

One point to consider this soiftware use share syncing may complicate the user interface during drive configuration.

It will require another field for disk in the user share configuration.

Then you will also have to contend with how split levels work and where the file is going to be copied to.

 

 

personally I don't like it, but wouldn't hurt to have it as an option

 

 

 

Need to be careful with how the delete function is implemented here.  A single delete could delete all 3/4 copies of critical files.

 

Well, RAID wasn't meant to prevent accidental erasing, so the user would just need to be careful like with any other drive.

Need to be careful with how the delete function is implemented here.  A single delete could delete all 3/4 copies of critical files.

 

Well, RAID wasn't meant to prevent accidental erasing, so the user would just need to be careful like with any other drive.

True,

Possibly more useful to some would be incremental backups to an alternate drive with HISTORY of the various versions of the file.

That way, an older version of the file, or a deleted file could still be recovered. 

 

There are many incremental backup schemes.  Over 20 years ago I had one incremental backup I wrote running hourly on a UNIX  machine where I had 20 developers on a project with a very tight deadline.  It saved our tails several times.  It used "find" with the "-newer" option and created a file with the current timestamp every time it ran to track files that changed since that time.

 

There are bound to be incremental backup packages available now that can be installed on unRaid.

 

Joe L.