Hi all,
I have the following issue in my production environment (solved by now using 1Gb cards and switch)
2 HP Proliant DL380P with HP 530T adapters
1 Netgear M7100-24X
1 EMC VNXe315 with 10GBe ports
The connection scheme is as following:
1 lacp lag between the 2 VNXe GBe ports and Netgear, working ok with 9000 MTU
1 10Gbe Proliant port connected to Netgear
I can mount the NFS datastore, I can ping (no networking issues at first sight) with 1500 packet size and 9000
If I start virtual machines on this datastore, or when I'm copying files to it, I get All Paths Down and datastore disconnects, recovering after a few minutes.
During this time, vmkping to the VNXe is working perfectly, without loosing packets.
I suspect of everything, from the switch to the NIC driver... any hints on this?
Update: I'm adding more information, this seems to work ok if I set the MTU back to 1500 (I've checked MTU was 9000 everywhere previously)
Can it be the switch? Anyone has tried those new 10gbe netgear?
Thank you very much in advance!