Page 1 of 1

R-Studio on recovering a ReFS volume on parity Storage Space

Posted: Wed Jun 03, 2015 1:02 pm
by berga_d
(This is a advanced technical question)
Before purchasing I need to know if R-Studio currently tries to recover corrupted ReFS metadata (and data too when Integrity Streams are enabled) from Parity and Mirror Slabs when deployed on these kinds of Storage Spaces.
Otherwise are you planning to add this feature soon?
Example: If R-Studio find that a ReFS cluster integrity checksum is wrong it should proceed to recover the cluster from the mirror copy in a mirror space, or to reconstruct it using parity data in a parity space.
Thanks for your software, you`re one of the very few vendors supporting both ReFS and Storage Spaces with a product reasonably priced!

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Wed Jun 03, 2015 2:29 pm
by Corsari
It is faster to download the shareware version of R-Studio and check directly on your drives.

Sorry but in this moment I'm not in front of my workstation so I can't check, it will be 10x faster to follow the suggestion above.

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Wed Jun 03, 2015 4:17 pm
by berga_d
I have already done that, and part of the files is corrupt/missing.
There is no way for me to discover if R-Studio is trying to use storage spaces parity "slabs" to find pristine copies of the corrupted clusters (and failing because both data and parity is corrupt),
or just ignoring them and reading only the data slabs.
Thanks for the fast answer.

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Wed Jun 03, 2015 10:37 pm
by Corsari
So it is better for you to wait for the official answer from Mr. Alt

At our company we use other processes with ReFS volumes

Are you in Europe?

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Thu Jun 04, 2015 2:57 am
by berga_d
Yes, I'll wait, I'm not in hurry.
(I sent you a PM)

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Thu Jun 04, 2015 11:01 am
by Alt
Actually, R-Studio recovers Storage Spaces and ReFS as two separate processes.
R-Studio uses mirroring and parity to recover physical sectors, and ReFS features to recover files.

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Sun Jun 07, 2015 1:18 pm
by berga_d
I supposed it was working that way.
The problem is that when you deal with corrupted data you have no way to know which of the parity (or mirror) slabs are the good/bad ones.
You can only find out if a slab is missing or not, and eventually reconstruct it.
This is the reason behind ReFS integrity streams: these checksums of data and metadata allow to find and fix corruptions (if other copies are available).
A simple example:

Mirror space
ReFS cluster checksum (integrity stream): 01

Slab 0 00110011 Computed checksum: 01
Slab 1 00110011 Computed checksum: 01

Then Somehting Very Wrong happens:
Slab 0 00110000 Computed checksum: 10
Slab 1 00110011 Computed checksum: 01

Through ReFS Integrity Streams you can now recover the good copy!

Re: R-Studio on recovering a ReFS volume on parity Storage S

Posted: Mon Jun 08, 2015 12:35 am
by Corsari
How did they corrupted the data?

Bad disks or user actions?