ririlot wrote:
I'm wondering if the upgrade from 4.0 has messed something up internally. When the host came back up, the upgrade had messed up all of the virtual switches and nic bindings. It had configured the management interface on one of the iSCSI ports for example.
It is possible of course that a inplace upgrade has messed up something with the vmkernel adapters. Strange, but possible.
You have doublechecked that no managment is enabled on the iSCSI VMK:s or vMotion or anything?
And that the management VMK does not have any iSCSI portbinding?
I can set up a session after reboot with "nc -z <ip of storage> 3260".
Could you also just check that it is possible to setup TCP through all interfaces / VMK interfaces?
http://rickardnobel.se/troubleshoot-iscsi-tcp-connectivity-from-esxi-with-netcat