Want to confirm some re-install points/

[lottathought]

Long story short ....kind of...

 

I recently upgraded my server. I am running Windows 8.1 Pro OS on a SSD. The only other things running are Drivepool and Scanner. The actual pool is a number of drives that are in addition to the SSD.

The OS SSD is not part of the pool but, as I said Drivepool and Scanner are installed there.

 

Soooooo..through work, I was offered the opportunity to use a copy of Windows Server 2012 R2 Essentials for my build and upgrade from the 8.1 P version. It will all be the same hardware that I recently upgraded to.

 

My issue is that I have moved my data over already...  All is stored and duplicated at the drive level in a Pool.

 

My planned process is....

Transfer out my licensees from Drivepool and Scanner.

Power down the PC and disco all the drives except the OS SSD.

Install Server 2012.

Install Drivepool and Scanner again...put the licenses back in.

Power down the PC and re-install all the drives.

 

It is my understanding at this point that Drivepool will find the earlier pool and re-create it...all files intact.

 

This leads me to my first question...how reliably does Drivepool accomplish this?   I ask because WHS Vs 1 made similar claims when the OS got corrupted a couple of years ago.    I reinstalled ....it was not pretty.

 

My next question is..... I have all files duplicated at the pool level.  When the system puts the pool back together and see's the files.....will it understand that the files are duplicated...or want to duplicate them again?   In other words...I already, through duplication, have 2 files on the drives. Will it recognize this...or will it want to make new duplications of the already duplicated files...giving me 4?

 

 

I think that I know the answers here...but since I am risking all my data, I want to make certain...hence the paranoia.

And of course..anybody that sees any flaws in what I am doing here, please point them out.

 

Many Thanks

 

 

 

 

[Christopher (Drashna)]

Ugh... I have a guide that I wrote for how to migrate the WHSv1 pool. Because I did it more than a few times.:

http://web.archive.org/web/20090411184621/http://wiki.wegotserved.com/index.php?title=Migrate_Shares_and_Backups_to_New_System

 

 

As for StableBit DrivePool... for testing I have 3 sets of virtual hard drives. Different sizes... I move them between VMs every time I test. And every time, it recognizes the pool on the "new" VM. I've also migrated to new hardware and reinstalled a few times, before coming to work for Covecube, and those went smoothly as well.

 

Specifically, we look for the PoolPart.xxxx folder and some hidden NTFS metadata to identify the pool, so there should be absolutely NO issue with moving the pool.

In fact, for the StableBit Server that I've been working on, I have a script that creates a pool automatically. And it works EVERY time.

It's that simple, and that effective. A lot of time and effort went into making sure the pool was SUPER simple to recover.

 

Additionally, as for duplication, we store the duplication status of the pool in the same sort of hidden NTFS metadata, so that this information is carried over with the pooled disks, as well.

 

Once the pool is recreated on the new system, DrivePool should initiate a measuring pass. This will check the status, and duplication or unduplicated files as needed (in case you changed anything, or are missing a disk, etc).

 

 

Additionally, worst case here, you would move all the data out of the hidden PoolPart.xxxx folders, delete those folders, add the disk to a pool and then "seed" the pool (move those files back). And this entire process would take maybe half an hour (at least for me).