0

Since 2012 Iam using the HP Proliant Dl385 G7 with Smart Array P410i.

  • 2xAMD Opteron 2.4GHz, 24 Cores
  • System Disk C: 2xSAS 15k in RAID1

Some time ago I purchased a new HP Proliant Dl380 G9 with Smart Array P440ar

  • 2xIntel Xeon 2.4GHz, 12 Cores, 24 Logical
  • System Disk C: 2xSAS 20k in RAID1

Both Systems running Win 2012 R2. Disks on both systems utilizing the transfer speed of 6Gbps. Only Hyper-V role is installed on both machines.

Some days ago I tried to compare the disk speed. Using IOMeter and CrystalDiskMark 6. Frustrating results brought me here.

Tests were performed during the lowest load (at least what I think).

Results from Gen7

enter image description here

Results from Gen9

enter image description here

My question is where from comes such a huge difference? What could I underestimate? Can it be that something is using the controller? Or maybe it is natural difference between these servers/controllers?

Dexterite
  • 170
  • 1
  • 6

1 Answers1

1

I'm not sure what's going on with your test, but there is no way that you're able to achieve a 5GBps throughput on a 2-disk SAS pair. You might be seeing something like a write-cache or ram disk that's operating that fast, or just a plain old bug in the software.

The output from the DL385 G7 is what I would expect to see for both systems. The P440ar might be a little quicker, but in almost every case, the disk itself is the limiting factor to the speed of reads or writes, not the controller.

Andrew
  • 2,142
  • 2
  • 19
  • 25
  • I played a bit with cache. You're right it is all because of it. Initially I thought I switched it off, but it looks like there are two separate triggers, need to read about them more carefully. Thank you. – Dexterite Feb 26 '18 at 21:40