Technical explanation

Configure a VSX LAG on both core peers and a matching LACP bundle on the access switch. The logical LAG can use available links to both core chassis simultaneously without ordinary spanning-tree blocking one of the uplinks. A peer failure leaves the remaining physical member operational provided the access switch and VLAN design are correct. Ensure consistent VLAN tagging, LACP parameters and an adequately designed VSX ISL/keepalive. Independent VLANs do not offer equivalent transparent redundancy, and static routes cannot replace a Layer 2 aggregated access uplink. With one member to each VSX core, an access switch can forward across both links. If one core fails, the LAG retains the surviving member instead of waiting for STP to unblock a redundant standalone trunk.

Putting each uplink into different access VLANs is segmentation, not a shared resilient bundle. Static routes to an access switch do not aggregate the pair’s L2 member links. A VSX LAG presents a logical LACP endpoint spanning the two cores.

Official HPE references | VSX and VSF Connectivity Design - Validated Solution Guide • Creating a VSX Pair | Study-guide topic: Managing Wired Networks with HPE Aruba Networking Central