P2000 G3 SAS Performance I/O in Question and Best Practices Using it in Hyper-V 2012 R2

[Shawn16]

Few part question here:

1.  Questionable I/O throughput with server and P2000, are my results on par with hardware/software setup or I have a configuration/hardware issue.  Are the 7.2K disks the limiting factor here?

2.  P2000 failover not working properly, preferred and current owner are always the same controller, which when a failover test occurs makes the failover take 5-7 seconds instead of instantaneous. Tech support said they should always be opposite and failover should be instantaneous.  Even when forcing controller to restart to make them say A/B or B/A shortly thereafter they report back to A/A or B/B.   

3. Best way to make these LUNs usable to Hyper-V VM's.  Will NOT be using as shared storage, just extra storage for VM's so that when performing a share nothing live migration of the OS, the P2000 storage can be accessed from all my Vms the same when migrating from one host to another.  I'm aware it can be done be pass through disk, VHDX (VHDX shared??) or possibly even mounting the disk in the VM, but which solution will do what i want and perform optimally.

Have a P2000 G3 SAS dual controller with latest firmware on controller with (24) 500Gb 7.2Krpm 2.5" 6Gbps drives with latest firmeware connected to a Dell R710 with 6Gbps SAS HBA.  Using dual mini-SAS cables, one to each controller for redundancy with MPIO on the server with Server 2012 R2.  MPIO config below. 

Question 1 Details

Have created several vdisks & volumes on the P2000 and then added them on the physical server as simple disks to test the difference in throughput, RAID0, 3disk RAID5, 6disk RAID5, all results basically the same.  Used CrystalDiskMark.  Basically the internal RAID1 OS and RAID5 Data storage have nearly twice the I/O as the external SAS P2000 storage and the internal disks are 3Gbps and the P200 are 6Gbps.

Controller Redundancy Mode - Active-Active ULP

Controller Redundancy Status - Redundant

Controller A Status - Operational

Controller B Status - Operational

vdisk - write cache policy (write-back), read ahead size (default) 

Volumes are mapped to all controller ports since  I plan on having multiple servers with access to same LUNs after sucessful testing is completed

 MPIO Configuration from 2012 R2 Server

MPIO Storage Snapshot on Thursday, 21 April 2016, at 14:20:56.862

Registered DSMs: 1
================
+--------------------------------|-------------------|----|----|----|---|-----+
|DSM Name                        |      Version      |PRP | RC | RI |PVP| PVE |
|--------------------------------|-------------------|----|----|----|---|-----|
|Microsoft DSM                   |006.0003.09600.18217|0020|0003|0001|030|False|
+--------------------------------|-------------------|----|----|----|---|-----+

Microsoft DSM
=============
MPIO Disk2: 02 Paths, Round Robin with Subset, Implicit Only
        SN: 600C0FF000140ABA63B5C35601000000                               
        Supported Load Balance Policies: FOO RRWS LQD WP LB

    Path ID          State              SCSI Address      Weight
    ---------------------------------------------------------------------------
    0000000077020004 Active/Unoptimized 002|000|004|003   0
        TPG_State: Active/Unoptimized, TPG_Id: 1, TP_Id: 7
        Adapter: Dell 6Gbps SAS HBA...                     (B|D|F: 005|000|000)
        Controller: 46616B65436F6E74726F6C6C6572 (State: Active)

    0000000077020003 Active/Optimized   002|000|003|003   0
      * TPG_State: Active/Optimized  , TPG_Id: 0, TP_Id: 1
        Adapter: Dell 6Gbps SAS HBA...                     (B|D|F: 005|000|000)
        Controller: 46616B65436F6E74726F6C6C6572 (State: Active)

MPIO Disk1: 02 Paths, Round Robin with Subset, Implicit Only
        SN: 600C0FF000140ABA24BEC35601000000                               
        Supported Load Balance Policies: FOO RRWS LQD WP LB

    Path ID          State              SCSI Address      Weight
    ---------------------------------------------------------------------------
    0000000077020004 Active/Unoptimized 002|000|004|002   0
        TPG_State: Active/Unoptimized, TPG_Id: 1, TP_Id: 7
        Adapter: Dell 6Gbps SAS HBA...                     (B|D|F: 005|000|000)
        Controller: 46616B65436F6E74726F6C6C6572 (State: Active)

    0000000077020003 Active/Optimized   002|000|003|002   0
      * TPG_State: Active/Optimized  , TPG_Id: 0, TP_Id: 1
        Adapter: Dell 6Gbps SAS HBA...                     (B|D|F: 005|000|000)
        Controller: 46616B65436F6E74726F6C6C6572 (State: Active)

MPIO Disk0: 02 Paths, Round Robin with Subset, Implicit Only
        SN: 600C0FF000140ABA7061955601000000                               
        Supported Load Balance Policies: FOO RRWS LQD WP LB

    Path ID          State              SCSI Address      Weight
    ---------------------------------------------------------------------------
    0000000077020004 Active/Optimized   002|000|004|001   0
      * TPG_State: Active/Optimized  , TPG_Id: 1, TP_Id: 7
        Adapter: Dell 6Gbps SAS HBA...                     (B|D|F: 005|000|000)
        Controller: 46616B65436F6E74726F6C6C6572 (State: Active)

    0000000077020003 Active/Unoptimized 002|000|003|001   0
        TPG_State: Active/Unoptimized, TPG_Id: 0, TP_Id: 1
        Adapter: Dell 6Gbps SAS HBA...                     (B|D|F: 005|000|000)
        Controller: 46616B65436F6E74726F6C6C6572 (State: Active)

MSDSM-wide default load balance policy: N\A

No target-level default load balance policies have been set.

================================================================================