We recently had a switch fail over that supports one of our iSCSI boot storage paths for ESXi VLAN6 from memory. As luck would have it most of our hosts lost access to their local boot storage and consequently changes could not be committed to the ESXi boot lun I image they had booted via VLAN6 at the time. My question is, should it have not have failed over to the alternate ESXi boot path or is this a supported function? I know "iSCSI multipathing is not supported at boot" but I assumed it would seamlessly failover in the event of a path failure once the OS was loaded and perhaps let me know via an amber alert. Just as a bit of background information we are using UCS kit with a vendor ESXi image and configured with an iSCSI boot dvSwitch exactly as per the vendor recommendations (see screenshot attached).
Of course it wasn't an end of days scenario regardless since once the network guys had resolved the problem I restarted the engagement agents and it seemed to reconnect again (presumably the original path).