Hello,
I am having this weird issue on one of the blades connected to a Nexus 1000v switch (version 4.0 00000) on a vSphere 4.1 environment.
There are more than 10 hosts connected to this virtual switch and everything works fine, but there is one single HP G7 blade that has problems. There are 3 uplink groups, management, vMotion and data. Each blade contributes with 2x1G NICs for management, 2x1G NICs for vMotion and 2x8G NICs for VM traffic (data).
If a VM is placed on a port group that uses the data uplink group it experiences packet loss; a continuous ping shows that every 2 or 3 seconds there is a packet dropped. Once one of the NICs on the uplink group is removed everything works fine and no packets are dropped, as soon as the uplink group has 2 NICs the problem comes back.
If the 2 NICs on the data uplink group are removed and added to a virtual standard switch with the default load balancing settings the problem never occurs.
The issue always happens with the same blade, if the same slot is populated with a different blade, it goes away.
Any ideas regarding this anomaly are highly appreciated,
Juan.
