Failing over 1xPS4000 and 1xPS6000 between 2x6224 stacks connected wth LAG to be able to do a fw upgrade

Hi,

I am about to failover the iSCSI traffic running between 6 Esx 4.0 hosts and 1xPS4000 and 1xPS6000.

My problem is that the Esx-version don't allow me to change iSCSI login timeout from 5 secs to 60 secs as recommended in the EQL docs. If this would work there wouldn't be a problem to do a controller failover but with this issue I am uncertain if I will lose iSCSI--connections causing some or all of the vm:s to go down when the passive controller becomes active in the "new stack".

I have tested to do a "manual failover" by taking a cable at a time from the active controller and connecting it to the "new stack" but this was done on 1xPS6100 and 1xPS4100 that didn't havie any iSCSI connections towards them.

My question is if this is a suitable option even when iSCSI-connections exist on the members or are there any better options?

Your advice is greatly appreciated

//Cris