Hi,

Are the servers connected directly to the MD or is there a switch in between? Losing the preferred path means that there is a loss of communication somewhere. It could be cabling a switch, updating firmware on both sides might also help. 

Does it happen to both controllers? When it happens are you able to ping the iSCSI host ports?

PING the iSCSI Host Ports.

1. Ping the iSCSI HOST
2. Open a command line.
3. Move to the directory containing SMI.exe (The default location is C:\Program Files\Dell|MD Storage Manager\client).
4. At the command line type ping [xxx.xxx.xxx.xxx] and press the key, where [xxx.xxx.xxx.xxx] is the IP address
   of one of the iSCSI controllers on the host. An unsuccessful ping returns the IP address.

Repeat step 3 with the address of the second iSCSI controller.

It's directly connected and the cabling has been replaced with brand new Cat 6 cabling and the SAN and hosts have all been firmware updated to the latest version but it still continues and occurs multiple times per day even when there is no load/throughput to speak of.

Two PowerEdge R720 servers

What model server is connected to it? 

Once it switches paths you can see if it can still ping the original path.

Wouldn't I have to do this at the exact moment the preferred path error occurs? That's going to be very difficult.

How are the servers directly connected? Is it two ports per server? Is there a reason why you are not using a switch? Can you provide a diagram of the connection? Is each host in its own subnet? Page 18 of the config guide. http://downloads.dell.com/manuals/common/powervault-md3200i_deployment%20guide_en-us.pdf

Well given I can immediately redistribute the disks -  I'd say a ping would work!

Both hosts are connected to each controller and both controllers error - it's not consistently controller 0 or 1 for example.

Ok, confirmed I've just had an alert for the preferred path error and both hosts can ping both controllers no issues.