nerdexam
HP

HPE7-A04 · Question #46

You want to configure your border leaf switches to leak routes between an overlay VRF named "dc1" and a VRF named "campus," which connects to the campus core. You also want to leak routes between an…

The correct answer is C. vrf dc1, import = 1:99, export = 1:1; vrf dc2, import = 1:99, export = 1:2; vrf campus, import = 1:1. Option C works because it uses a hub-and-spoke route-target design where each DC VRF has a unique export RT, while the campus VRF acts as the hub with a single shared export RT (1:99). dc1 exports 1:1 (campus imports it), dc2 exports 1:2 (campus imports it), and campus exports…

Designing Data Center Network Architectures

Question

You want to configure your border leaf switches to leak routes between an overlay VRF named "dc1" and a VRF named "campus," which connects to the campus core. You also want to leak routes between an overlay VRF named "dc2" and the campus VRF. "dc1" should not be able to communicate with "dc2." Which is a valid design for the VRFs' IPv4 route-targets?

Options

  • Avrf dc1, import = 1:1, export = 1:99; vrf dc2, import = 1:2, export = 1:99; vrf campus, import = 1:99,
  • Bvrf dc1, import = 1:1, export = 1:1; vrf dc2, import = 1:2, export = 1:2; vrf campus, import = 1:1, 1:2,
  • Cvrf dc1, import = 1:99, export = 1:1; vrf dc2, import = 1:99, export = 1:2; vrf campus, import = 1:1,
  • Dvrf dc1, import = 1:1, export = 1:1; vrf dc2, import = 1:1, export = 1:1; vrf campus, import = 1:1,

How the community answered

(42 responses)
  • A
    14% (6)
  • B
    24% (10)
  • C
    57% (24)
  • D
    5% (2)

Explanation

Option C works because it uses a hub-and-spoke route-target design where each DC VRF has a unique export RT, while the campus VRF acts as the hub with a single shared export RT (1:99). dc1 exports 1:1 (campus imports it), dc2 exports 1:2 (campus imports it), and campus exports 1:99 - which both dc1 and dc2 import - creating bidirectional reachability to campus while keeping the DCs isolated from each other, since neither dc1 nor dc2 imports the other's unique export RT.

Option A breaks return path connectivity: dc1 and dc2 both export 1:99 and campus imports 1:99, so campus receives routes from both - but campus has no export RT that matches dc1's import (1:1) or dc2's import (1:2), so no routes can flow back from campus to the DCs.

Option B violates the isolation requirement: if campus imports both 1:1 and 1:2 and then exports those learned routes back with RTs that dc1 and dc2 import, dc1's routes transit through campus and leak into dc2 (and vice versa), breaking the dc1 ↔ dc2 isolation.

Option D is immediately disqualifying: dc1 and dc2 both use identical export and import RTs (1:1), so they directly import each other's routes and can communicate freely.

Memory tip: Think of option C as a "one-way funnel to campus" - each DC pours its routes into a unique pipe (1:1 or 1:2), but both DCs only hold the same key (1:99) to receive campus routes back. Since neither DC holds the other's key, they can't open each other's pipe.

Topics

#VRF route leaking#BGP route-targets#VRF isolation#EVPN overlay

Community Discussion

No community discussion yet for this question.

Full HPE7-A04 Practice