Storage Spaces vs Plain-Old-Partitions on single-spindle 300GB drive

When I originally installed my first WS2012 system (running WSUS, DHCP, WDS), I created simple volumes on the drive as I've always done. OS is installed on a 150GB drive, and I have a second 300GB drive that's providing storage for WSUS, WDS, and who knows what else I'll do with the 70% of that drive that WSUS/WDS will never use.

I'm familiar with the concepts and intents for Storage Spaces, but is there any compelling reason to use Storage Spaces on a single-spindle drive? If not, then I'm happy to just leave it as it is and invest my time in other ende

December 18th, 2013 2:39pm

When I originally installed my first WS2012 system (running WSUS, DHCP, WDS), I created simple volumes on the drive as I've always done. OS is installed on a 150GB drive, and I have a second 300GB drive that's providing storage for WSUS, WDS, and who knows what else I'll do with the 70% of that drive that WSUS/WDS will never use.

I'm familiar with the concepts and intents for Storage Spaces, but is there any compelling reason to use Storage Spaces on a single-spindle drive? If not, then I'm happy to just leave it as it is and invest my time in other endeavors.

If you'd create Spaces with redundancy (mirror or parity) and use ReFS on top you'll have much more stable config. Classic LVD + NTFS are not capable of handling anything related to silent data corruption (if SCSIOP_READ10 completed OK storage stack will think read is OK and NTFS has no mechanism to check for data integrity either). SS + ReFS will both detect and try to fix the error a) mapping new block and b) relocating data using alive content. That's in theory :)


Free Windows Admin Tool Kit Click here and download it now
December 18th, 2013 4:59pm

If you'd create Spaces with redundancy (mirror or parity) and use ReFS on top you'll have much more stable config. Classic LVD + NTFS are not capable of handling anything related to silent data corruption (if SCSIOP_READ10 completed OK storage stack will think read is OK and NTFS has no mechanism to check for data integrity either). SS + ReFS will both detect and try to fix the error a) mapping new block and b) relocating data using alive content. That's in theory :)

Okay.. but is it worth the effort for a single-spindle DAS drive?
December 18th, 2013 10:42pm

[ ... ]

Okay.. but is it worth the effort for a single-spindle DAS drive?

It does not make any sense to wrap extra layers above hardware to increase storage stack depth in this case. It would lead to higher latency and lower performance. + add some code could potentially fail.


Free Windows Admin Tool Kit Click here and download it now
December 19th, 2013 5:50am

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics