A couple of my Colleague, Claus Andersen and Jørgen Rosenkvist Pedersen have had some customers that have had problem with Citrix NetScaler VPX and HPE Blade C7000 with virtual connects. The problem may also exist in other types of hardware
After vMotion of the Citrix NetScaler VPX from one blade to another, one of the network ports in the destination ESXi host stops forwarding packages, but the port does not shutdown. Some time it happens at ones og after a shot time. Virtual Connect rapports a “Loop Detection”.
Citrix is aware of this problem and has updated there FAQ on the Citrix Netscaler VPX with information, about this problem and “Solution”: Citirx FAQ
Can XenServer XenMotion or VMWare VMotion be used to move active instances of NetScaler VPX? NetScaler VPX does not support XenMotion or VMotion. XenMotion and VMotion support are being evaluated.
So the solution is to lock the Citrix NetScaler VPX to a host, and not move it online.
I Hope Citrix will find a better solution
Update: Citrix now supports vMotion, from version 12.1, see Citrix Knowledge Base article here: https://support.citrix.com/article/CTX121291