Refer to the diagram below.

The correct answer is D. Allowed VLANs between access switches. The diagram shows Server-1 connected to Access-2 in VLAN 20 and PC-1 connected to Access-1. The inter-switch link between Access-1 and Access-2 is shown as a trunk carrying VLANs 1, 10, and 20. If users on Access-1 cannot reach Server-1 in VLAN 20 after a recent change, while devices directly connected to Access-2 are not affected, the most likely issue is that VLAN 20 is not being correctly carried across the trunk between the access switches. HPE Aruba Networking AOS-CX documentation states that vlan trunk allowed defines which VLAN traffic is allowed across a trunk interface. It also explains that allowed VLANs are the VLANs that can be transported by the trunk; if a required VLAN is missing from the allowed list, that VLAN’s traffic will not traverse the trunk. Option A is less likely because devices directly connected to Access-2 can access Server-1, indicating Server-1’s local VLAN placement is probably correct. Option B is unlikely because a down trunk would affect more VLANs than just access to VLAN 20. Option C is not supported by the symptom pattern.