10GbE iSCSI with Midplane 1.0

Hello,

I have a question about using KR-based Mezzcards and IO-Modules in a 2010' M1000e enclosure using midplane 1.0.

Short intro:
We are running 5x M610 and 2x M620 blades in a M1000e enclosure. Each running Hyper-V serving 4-6 VMs per node. Each node is connected

• using Passthrough Module in Fabric A to network
  and
• using M6220 in Fabric B to a M3000i storage.

Since I/O looks to be a bottle-neck in that setup, we are planning to go to 10GbE iSCSI. I talked to our sales rep. which recommends buying new enclosure, new switches, new passthrough, new storage, ... - so nearly everything --- only because Midplane 1.0 is not capable of 10GbE. Even if the new enclosure is not that expensive, there is some effort needed on our side to replace it.

However in the "PowerEdge M-Series Blade I/O Guide" I can read on page 89:

Q: Can I use KR-based mezz cards and switches on fabrics B and C of my existing chassis?
A: Yes, fabrics B & C fully support 10GbE KR on any midplane.

So instead of replacing the whole M1000e and all IOM, I am thinking of the following:

• PS4210X or MD3820i storage
• PowerConnect M8024-k 10GbE Switch for Fabric C (2x)
  -or-
  10Gb Ethernet Pass Through -k for Fabric C (2x) and an some external 10GbE switches (i.e. N4032)
• Broadcom 57712-k Dual Port 10GbE NDC for Fabric C of each M610/M620 (so 7x)
• some SFP+ cables

So question is: Will M8024-k or the 10GbE-k pass through module work in M1000e on midplane 1.0 on Fabric B/C? And if yes, is it recommended, i.e. will there be any performance issues (will it run on 10GbE) or something else to keep in mind? Did is missed something?

The advantage of this setup would be that the "old" 1GbE iSCSI cluster will just continue working fine and I could migrate the VMs step by step to the new 10GbE storage. This disadvantage is that I continue running an old/outdated enclosure, for which I have to spend some money for extending the support.

Thanks in advance!

BR,
Markus