Question 760
For a regional peering-exchange operator, the change record describes: An enterprise customer is promised IPv6 VPN reachability but the PE provisions only VPNv4 MP-BGP and v4 VRF interfaces. What preventive validation should be standard?
Explanation
Correct Answer
Validate VPNv4 and VPNv6 service route counts independently. This helps prevent recurrence of the issue: An enterprise customer is promised IPv6 VPN reachability but the PE provisions only VPNv4 MP-BGP and v4 VRF interfaces. Acceptance evidence should be: An IPv6 test prefix reaches the remote customer site without unintended mixing with IPv4-only service routing.
Incorrect Answers
Validate CE route reception, VPN export and remote import as three separate checks. That is a sensible control for ANOTHER scenario: A CE-PE BGP session successfully exchanges routes but a newly provisioned customer VRF has no route targets and none of its routes propagate to distant PEs. It is less targeted to this issue than: Validate VPNv4 and VPNv6 service route counts independently.
Require overlapping-address negative tests before new VPN attachments. That is a sensible control for ANOTHER scenario: Two customers use the same 10.10.0.0/16 prefix but a PE leaks VPNv4 routes into its global table, incorrectly merging their routing contexts. It is less targeted to this issue than: Validate VPNv4 and VPNv6 service route counts independently.
Review RPL ownership and ring state during ring expansion. That is a sensible control for ANOTHER scenario: A G.8032 Ethernet ring sees a broadcast storm after a ring protection link is accidentally forwarding on both sides of a stable ring. It is less targeted to this issue than: Validate VPNv4 and VPNv6 service route counts independently.