Hi GreedDx,

Well, it seems EMC gives Contradictory information when it comes to HA and MPIO/Jumbo Frame configuration. In pure VMWare Native language, MPIO Means Multipathing, Jumbo Frame, Means MPIO along with Multipathing and Port Binding.

I have setup my VNXe3150 with two Subnet VMware Communities: Hussain Al Sayed's Blog: Setting up & Configuring EMC VNXe3150 iSCSI SAN Storage with High Availabil…

SPA-Eth2 10.90.8.1

SPB-Eth2 10.90.8.2

SPA-Eth3 10.100.8.1

SPB-Eth3 10.100.8.2

Now, as per VMWare Support Port Bindings is not supported configuration when it comes to different broadcast domain. And VMware supports Port Binding only on single broadcast domain.

EMC with their HA documentation along with the domain made by Matthew Brender & Henri Hamalainen in the Ask the Expert of "VNXe Front-End Networks with VMWare" gives contradictory information both of the document speaks about HA with different subnets and Ask the Expert document clearly elaborating on the port binding, Round Robin and changing the default IOPS to gain high storage performance.

Now which is supported configuration, I don't and I left it to EMC to answer me on that.

Thanks,

Hi greendx

You have 2 dual port cards, one on each SP. That means you can have at the most 2 subnets connected to those. You cannot connect each of these four ports  to different subnets.

From page 11 (Fail-safe networking), "To take advantage of FSN, you must cable both SPs the same way." That means, in this case, if you are connecting SPAP1 to switch A, then SPBP1 also should be connected to switch A.

Again, as far as networking goes, it collectively calls SPAP1 and SPBP1 as one say eth10. So from your vmware side, it doesnt see it as 4 nics, it sees as 2 nics.

For your second question also, my response above applies. So, with these 2 port cards, you will be able to just setup one LACP trunk, which would be eth10 and eth11. It internally does the needed trunks of SPAP1+SPAP2 and SPBP1+SPAP2. See Page 18.

Thanks.

Ok yes vmware part makes sense, I confused my self with that.

As for the network config basically you're saying trunk both ports on controller A in the VNX (each being assigned an IP on a different subnet) and that would give that eth10 and do the same on controller B. Then LACP trunk them all on the network side. Makes sense. It's just not at all clear in the documentation. I even ran it by our EMC guy and he said not to trunk the ports on the VNX and do 2 separate LACP trunks on the network side.

So this the config is what we ended up with. However, when carving out LUNs and presenting to vmware I'm only able to see 1 path to each LUN.

SPA-eth10 - 10.10.2.2
SPA-eth11 - 10.10.3.2

SPB-eth10 - 10.10.2.3
SPB-eth11 - 10.10.3.3

SPA-eth10 - Switch01-P02
SPA-eth11 - Switch02-P02

SPB-eth10 - Switch01-P03
SPB-eth11 - Switch02-P03

vmnic3 (Onboard NIC Card) - Switch01-P04
vmnic6 (External NIC Card #1) - Switch01-P05
vmnic8 (External NIC Card #2) - Switch02-P04
vmnic9 (External NIC Card #2) - Switch02-P05

vSwitch2
iSCSI01a - vmk3 - vmnic3 - 10.10.2.4 (active) // vmnic9 - 10.10.2.5 (unused)
iSCSI01b - vmk4 - vmnic9 - 10.10.2.5 (active) // vmnic3 - 10.10.2.4 (unused)

vSwitch3
iSCSI02a - vmk5 - vmnic6 - 10.10.3.4 (active) // vmnic8 - 10.10.3.5 (unused)
iSCSI02b - vmk6 - vmnic8 - 10.10.3.5 (active) // vmnic6 - 10.10.3.4 (unused)

Beaware that using port binding with Different subnet its not supported by

VMware.

Where does it say that? I'm pretty sure that this is fairly standard config from vmware side.

Thanks for the link to that first KB. I just removed the iscsi bindings. However, now I'm no longer able to see any storage paths

This another article from VMware explaining issues when rescaning vmhba

"iscsi adapter" it takes a bit longer when using MPIO, active/unused NICs

and port binding.. I have faced the three symptoms outlined in the first KB

articles with VNXe3150 on one broadcast domain. Slow scan, vmKernel logs

full of SCSI failures and when I test SP failover by powering SPa off, VMS

stopped responding and ESX timed out accessing the LUN that was owned by

SPa. When LUN transferred to SPb, connecting back on ESX and VMs started

pinging again.

I have left my configuration with more than one broadcast domain,

MPIO/active/unused NICs, and port binding of two vmKernels portgroups. All

issues sorted out.

Today EMC high level technical enginner confirmed this configuration is

good and no point of changing it since there it is working perfect.

http://kb.vmware.com/selfservice/microsites/search.do?cmd=displayKC&docType=kc&docTypeID=DT_KB_1_1&externalId=2017084

Here VMware KB which explains this; there is another article but can't find

it. http://kb.vmware.com/kb/2038869

Yes that for sure.. I faced this issue with the VNXe3150 too, I see only

two paths even with ports bindings. Here how I set it up:

http://communities.vmware.com/people/habibalby/blog/2013/03/12/vnxe3150