X1600 12TB Raw Space vs. Available Space

[Aaron Schneider]

I have 12 1TB SATA drives in an X1600. By default the disks were divided into two arrays as following

Array A contains:

Logical Drive 1 (40GB for the OS)
Logical Drive 2 (5GB for Recovery)
1.7 TB of UnUsed Space

Array B

All remaining UnUsed Space (~9+TB)

This was verified using the ACU.

I've begun creating shared folders with the Automated Storage Manager and have utilized 95% of the UnUsed space on Array B. All my volumes are RAID 0 with NO Snapshots, so there is virtually no overhead.

The ACU and Storage Utilization view of the ASM both correctly show that I have almost ~2 TB of UnUsed space still sitting on Array A. However I am only able to allocate space from Array B when I create folders. I realize the difference between RAW storage and available user storage, but with no RAID/SnapShot overhead, I'm trying to understand what this nearly 2TB of Unused space on Array A is there for?

I opened a case with HP regarding this issue and was told that this is by design and that that space was unavailable to allocate. However the technician could not elaborate on what that design was or what was actually occupying the space or why it wasn't allocatable. In addition, I was told that manually allocating a volume from the Unused space on Array A was not adviseable....but again, could not be told why.

I see nothing in any HP documentation indicating there is a disk requirement of any kind for the X1600 that will, by default, reduce the capacity of the unit by almost 15%.

I have an additional 500GB to allocate to the X1600 and need a portion of that ~2 free TB. Can anyone confirm whether or not this space is in fact useable...either from the ASM or via the ACU/Windows Storage Management...and if not, point me towards any documentation that would explain why?

Thanks...

[TTr · Accepted solution]

Array A uses 2 disks (~2TB) and it is used for the OS exclusively.

Array B uses the 10 disks and 9+ TB of space.

The ASM does not allow you to allocate space from array space because it would interfere with the i/o from the OS volume.

If you can bypass ASM and allocate volumes from array A, go ahead and do that but keep in mind that any i/o from these additional array A volumes would come from the same two disks with the OS volume. Hense OS i/o impact.

[Aaron Schneider]

Ok, that makes sense. Since the OS is going to be hosted on its own array and is at least RAID 1, it will require at least 2 of the disks. Since my disks are 1 TB, I am unfortunately left with a lot of space on that array that the ASM cannot allocate from. But like you hinted to, if I allocate the space manually and use it for something like backup data that needs to be accessed quicker than tape but whose day-to-day I/O is low since its stagnant data, any contention should be minimal?

Thanks...

[TTr]

Yes if you put stagnant or minimally accessed data on additional volumes in array A there would be no interference with OS i/o. In general OS i/o is minimal to begin with.