0

I adopted a virtual Windows Server 2012 Storage Server that has 4 QNAPs attached via iSCSI.

The storage environment is connected via iSCSI through a Cisco SG300 (Gigabit); jumbo frames is not configured.

Each QNAP is individually configured as a RAID 6 with 8 drives each unit.

(2) of the QNAPs are configured in the Storage Server as a MIRROR via Disk Management

(2) of the QNAPs are configured in the Storage Server as a MIRROR via File and Storage Services using Storage Pools

For whatever reason, the QNAP mirror via Disk Management works as expected and the QNAP mirror via Storage Services performs poorly.

Bandwidth wise, I'm expecting around 60-100MB/s, but I'm averaging 30MB/s. I do get a spike every now and again around 80MB/s, but I also get long bouts of no throughput whatsoever. It's not consistent.

CPU utilization is under 5% and memory is steady at 21%.

Is this discrepancy in my expectation due to the Storage Server mirroring process or is this expected of RAID6 in general? I'm running perfmon now to collect data, but it seems odd that there'd be such poor and inconsistent performance.

UPDATE:

PS C:\Windows\system32> fsutil fsinfo ntfsinfo d:
NTFS Volume Serial Number :       0xb0fe9f7ffe9f3c92
NTFS Version   :                  3.1
LFS Version    :                  2.0
Number Sectors :                  0x000000077ffbefff
Total Clusters :                  0x00000000efff7dff
Free Clusters  :                  0x000000009991ea4a
Total Reserved :                  0x0000000000000000
Bytes Per Sector  :               512
Bytes Per Physical Sector :       4096
Bytes Per Cluster :               4096
Bytes Per FileRecord Segment    : 1024
Clusters Per FileRecord Segment : 0
Mft Valid Data Length :           0x0000000000040000
Mft Start Lcn  :                  0x00000000000c0000
Mft2 Start Lcn :                  0x0000000000000002
Mft Zone Start :                  0x00000000000c0040
Mft Zone End   :                  0x00000000000cc840
Resource Manager Identifier :     F19B79FA-AFE8-11E5-80C0-3CA82A1CECDB

PS C:\Windows\system32> fsutil fsinfo ntfsinfo f:
NTFS Volume Serial Number :       0xa832ab9d32ab6f4e
NTFS Version   :                  3.1
LFS Version    :                  2.0
Number Sectors :                  0x0000000ffffbefff
Total Clusters :                  0x00000000ffffbeff
Free Clusters  :                  0x0000000095e4dba5
Total Reserved :                  0x0000000000000000
Bytes Per Sector  :               512
Bytes Per Physical Sector :       512
Bytes Per Cluster :               8192
Bytes Per FileRecord Segment    : 1024
Clusters Per FileRecord Segment : 0
Mft Valid Data Length :           0x0000000000040000
Mft Start Lcn  :                  0x0000000000060000
Mft2 Start Lcn :                  0x0000000000000001
Mft Zone Start :                  0x0000000000060000
Mft Zone End   :                  0x0000000000066420
Resource Manager Identifier :     E628C5C7-CA70-11E5-80C5-3CA82A1CECDB
CIA
  • 1,604
  • 2
  • 13
  • 32
  • I can't find it now but I'm sure I read that the disk allocation unit (default) value differs between Disk Management and Storage spaces. Type fsutil fsinfo ntfsinfo and see if they are the same, if not this could be a reason – Drifter104 Feb 10 '16 at 17:49
  • Interestingly, `bytes per physical sector` and `bytes per cluster` differ as you indicated. I don't suppose there's a way to change those values in Storage Spaces (because the Disk Management's config seems to provided better throughput). – CIA Feb 10 '16 at 19:13

0 Answers0