Hi Don,

I always appreciate reading the answers you provide on various posts here, but, I do need to question your statement "If you have set the servers to best practices, they should be able to ride out the failover restarts.  I would do this during a low I/O period to minimize the failover time."

Our servers are configured to best practices and in multiple firmware upgrades, many of our servers were not able to ride out the failover restarts.  By extension, this means our servers would not be able to ride out an unexpected controller failure, since that would incur a similar failover between controllers.

I worked at great length with the Level 2 & 3 EqualLogic engineers in Nashua to identify our issues, but, they could not find any problems.  They agreed our entire stack from servers to switches to arrays are configured per best practices.  And for a customer with a modest SAN, there truly is no low I/O period and so the engineers state what this really means is a complete "lights out" maintenance window to do SAN firmware updates.  I'm afraid this message isn't properly reaching customers and while I've no doubt your statement was made in good faith with the best knowledge available, the reality is much more of a roll of the dice when performing firmware upgrades.

Thanks,

Ryan