Performance-Problems with NetApp Storage System and HP NIC-Teaming

At the moment we are implementing a new NetApp Storage-System (with two controllers) at our Company and had in interesting error when we tried to configure the NetApp-Controllers via http. A connection to both NetApp-Controllers using http worked without any error from a Client with a single NIC. When we tried to configure the NetApp-Controllers from a HP-Server with teamed Network-Cards one controller was reacting normally and one was really slow. The same effect happened on SSH-Connections. When you pressed several times the browser-refresh-button on the normal working controller it started to react really slow – but then the other controller was working normally without any delay. After some searching at Google I found an article which mentioned to run the following command on every NetApp-Controller:

options ip.fastpath.enable off

After running this command on both NetApp-Controllers it was possible to manage both NetApp-Controllers without any delay – even from HP Servers with teamed NICs. So the interesting question was: What happens if this option is disabled? So we started to search in the NetApp-KB and found the KB-Article-ID 2013130.
In this Article NetApp describes exactly the errors we had – and the given solution is the above mentioned option.
But the coolest thing is the Cause for this:
”The cause for this issue is unknown”.

So if you are using HP-Servers with teamed NICs and a NetApp-Storage-System just disable the fastpath option and everything is going to work really fine

PS:
The fastpath option is described in the manual – it caches Mac-Addresses.