Trying to get the wording right because we keep coming back to traditional unRAID parity solutions which might not be the only way to do this:

 

"Protected Disk: A disk whos contents can be restored using any mechanism"

 

To be restored it must have been stored.  I'm assuming this includes the worst case where the disk is lost or not functioning.  Describe the location of the data that is to be restored now that the drive is dead.  How was the data stored while offline?

 

or are you actually asking for ideas on how to store hard drive contents in some non-hard drive media...tape backup for instance?

 

 

BTW, in the spirit of recycle-reuse you should sell those small older hard drives on ebay as fast as you can and move the contents to a single large drive if possible.  

Yes this is a for fun thread to throw about ideas of how it could be done.

 

Just an FYI I have never sold, returned or disposed of a HDD ever. At EOL they are securely stored ad nausium. But thats OT

Yes this is a for fun thread to throw about ideas of how it could be done.

 

Just an FYI I have never sold, returned or disposed of a HDD ever. At EOL they are securely stored ad nausium. But thats OT

 

Data management is so un-fun.  Most home users don't have the time, interest nor the equipment to deal with it.  Business people have the mandate to protect information so they invest in manpower and equipment to ensure the data is protected and restorable.  Tape backup is cheaper than storing hard drives.  For the soho user keeping the data online or a couple extra hard drives usually does the job.  Home users want to be, well users and not administrators.  

 

For most of my drives that aren't part of my unRAID array and I use SNAP to access, aren't in the array because I don't want to spend the time to do even that much management.  Some of the data (drives) changes and I just want to be able to get at it if needed.  Heck, I have automated the ripping of DVD's because I don't want to spend time doing that either.  As much as possible I want to be a consumer or user of the data and spend as little time as possible handling it or maintaining it.  

 

What is fun is when getting the data into the array is made easier, simpler without my oversight.  Rules based, smart device actions.  I want my computers to work for me and not the other way around.  

So what I am hearing is that it has to be simple. I would definitely agree and go one further and say in a trade off between simplicity and efficiency then simplicity should win hands down.

 

How about this then. SNAP is nice and simple. How about a SNAP like application where you tell uNRAID about a bunch of disks. In unRAID GUI you pair disks into groups of 2. Disk 1 being data and Disk 2 being backup.

 

When you plugin disk 1 unRAID prompts you to plugin disk 2 and it rsyncs them. From now on unRAID remembers the amount of changes to disk 1 and presents via the GUI a representative % of how much has changed between disk 1 as of now and your last backup.

 

This is protection like WHS but would allow an infinite number of drives and only two sata docks.

 

Far less efficient than parity but far far simpler. Allows you utilise a SNAP like application for ease but afford you protection with little complication. If you treat the data drive like a closed off collection that doesn't change you would never need to ever plug in disk 2 again until disk 1 broke.

 

Think about it

Do you want fault tolerance or do you want a backup?  Trying to do both is the problem.  I see more problems with systems when people try to stretch the system's envelope to make it fit something not originally intended.  I was brought in on a contract once at Ft. Bragg, by a company that had tried to adapt iits maintenance system that ran on an AS/400 to handle housing maintenance at the base..... the only problem was their home-grown maintenance software was designed for aircraft maintenance.

 

unRAID is not backup... it is fault tolerance with a low parity:data ratio and high survivability (i.e. data from non-failed drives is always recoverable).  unRAID has a capacity of XX drives.  Pushing the envelope with edge cases of marginal use to the customers in the center of the bell curve, is not where development man-hours are rationally spent.

Surely many uRAID user have retired disks?. We recommend it all the time.

 

Does it matter what you call it the end goal is the same.. surviving a dead disk.?

 

But if its inclusion is a stumbling block ignore unRAID completely. This is a for fun thread anyway... its more about the theory than the practice.

Pushing the envelope with edge cases of marginal use to the customers in the center of the bell curve, is not where development man-hours are rationally spent.

 

I'm still decoding that sentence.  Which part refers to you and your beliefs and which part refers to people you think are wasting time?

The OP says it's a brain excercise so let it happen.  It's still in the analysis stage. 

Surely many uRAID user have retired disks?. We recommend it all the time.

 

Does it matter what you call it the end goal is the same.. surviving a dead disk.?

 

But if its inclusion is a stumbling block ignore unRAID completely. This is a for fun thread anyway... its more about the theory than the practice.

 

Are you trying to find a creative use for old hard drives or do you just want to protect data that is offline?  

 

One of the types of data I keep that I don't particularly need backed up is an old or prior OS.  I have had lots of machines and at some point they are superceded or consolidated onto new equipment.  If that means a new install then I take the old hard drive with everything exactly as it was when running and put it on a shelf.  I do that because there may be stuff I need off it but I'm too busy getting the new machine(s) up and running and configured.  Well, now in addition to just the shelf, I can plop it into my dock where it's very simple to get to.  I don't really want to dump all that data into my array.  I just want it available, easily and quickly.  SNAP makes it easy to mount/share it.  Because the sharename represents WHAT the disk has on it then just finding it in the right place becomes easy and reliable and I know I'm getting to the data I intended to.  

 

People think compartmentally.  If you have a remote for each device you would be shocked if one remote unintentionally caused changes on a different device.  Yet use a universal remote and the compartment now includes all the devices as long as you press the device selection button first.  That's the same as having several remotes because the device selection button changes the remote's functionality in our minds.

In automating my DVD ripping I bought 4 DVD drives.  One for my daughter's games.  One for general use - no assignment.  Two are for ripping only, one for Main movie and the other for Full DVD.  They all can do the same thing.  So what's the difference?  Why not just have one?  Because the hardware (position, color, name, etc) defines the actions that will be performed.  I have software running so just placing a disc in the tray starts the ripping. 

 

For some of my hardware, I may only want a small subset of it's total capabilities.

I should be able to decide how I want to use my hardware and once decided I want a particular type of actions performed I shouldn't have to decide that again every time I choose to use that hardware. 

Let me extend that into another device, my camera.  I have a place in my shares for pictures.  When I plug my camera's usb port into my unRAID system it connects, mounts and sync's the pictures to where I expect them to be and then unmounts and beeps so I can pull my camera off the port.  Since I always want that action when I plug that hardware in why not automate it?  That's one of SNAP's uses.  It's making the machine smarter.  My fun-to-work ratio goes up and I'm being a user instead of an admin.