We need to perform firmware updates to our Equallogic SANs. They are on a dual-controller where it first would update the stand-by then failover and update the other controller. However, this process still interrupts iSCSI connections for a brief period of time. Our VMs are mostly Windows and the OS can handle up to a minute of disconnection to the back-end SAN, but we also have many SQL workloads and SQL does not handle this well at all (we saw some corruption the last time we updated our SANs)
So we're considering this time around maybe "Saving" or "Pausing" all the VM's first, performing the SAN firmware updates, then resuming all the machines in Hyper-V. I'm not really sure what is best however (saving or pausing), or what other alternatives to consider. I know pausing doesn't save the state anywhere which means if the node happens to bluescreen or reboot for some reason the state of all those machines on that node would be lost.
Saving seems to make more sense but wanted to get a feel for how others are handling updates like this. Any advice would be appreciated here.
One cluster runs Server 2012 R2 node, and a separate cluster runs Server 2012 nodes (different SANs for each). The VMs on the nodes are mostly Windows machines, but may be running anything from Server 2003 to 2012 (most are running Server 2008 R2). SQL would also be a variety of versions from SQL 2005 to 2012, most being SQL 2008 or SQL 2008 R2.