4200, ESXi & SFP+ Problems

[DanW85]

Hi there,

after I ended up purchasing Cisco SFP+ cables (CISCO SFP-H10GB-CU1M= if anyone's intersted) I seem to be running into some problems with iSCSI targets. The cables seems to be working absolutely fine, green lights at both ends of the cables.

Set-up:
ReadyNAS4200
Target name: vmware1
LUN Description: vmware-pool-1
CHAP Authentication = off
Access control = on
This is where I though there might be a problem, but I have copied and pasted the IQNs straight from VMWare.

4200 networking:
I have tried both the following

eth3: 172.31.10.1
eth4: 172.31.10.2

eth3: 172.31.10.1
eth4: 172.31.10.1
at this point the server who could actively see the iSCSI target changed to the opposite host(!?)

VMWareHost1
Network Adaptors: vmnic4, vmnic5
Storage Adaptors: Dynamic Discovery: Send targets ==> 172.31.10.1, 172.31.10.2

VMWareHost2
Network Adaptors: vmnic4, vmnic5
Storage Adaptors: Dynamic Discovery: Send targets ==> 172.31.10.1, 172.31.10.2

I've raised a support ticket with VMWare but I've a feeling it's a combination of something I'm doing wrong on the 4200.

What do I need to do differently? Is the subnet mask wrong or something silly like that?
I should note that the 172.31.10.x network is only available on the readynas4200 via that cabling, and the VMWare hosts.

[DanW85]

OK, so it looks like it's not a problem with anything on the VMWare set-up side of things.

I only seem to be able to ping one interface at a time, and it sometimes changes which port this is (so it doesn't look like a faulty port on the ReadyNAS's SFP+ card...yet)

Are there any limitations on how this direct attach can be cabled? i.e. can it only handle one host at a time?

[DanW85]

Turns out it was some sort of routing problem with the ReadyNAS.

eth3 is now on 172.31.11.x and eth4 is on 172.31.10.x