How should the architect configure high availability (HA) on a XenServer host to support an iSCSI SAN?

Simply connecting a minimum of two network interface cards without ensuring they are on separate network paths does not fully address high availability, as a single switch or network segment failure could still disrupt storage access.

Simply connecting a minimum of two network interface cards without ensuring they are on separate network paths does not fully address high availability, as a single switch or network segment failure could still disrupt storage access.

Connecting a minimum of two network interface cards 'each on a different' without specifying 'separate physical path/switch' might imply logical difference, but for robust HA, physical separation (as implied by 'separate' in D) is essential to eliminate single points of failure.

To achieve robust high availability for iSCSI SAN connectivity on a XenServer host, it is crucial to connect a minimum of two network interface cards, with each card configured on a separate physical network path (e.g., different switches), to eliminate single points of failure and enable multipathing.