Advanced BGP Practice Questions

Answers and explanations appear when you work through the set above.

1. Weight ties between two iBGP-learned paths. Path A carries local preference 150 and Path B carries local preference 100 (the default). Which path becomes the best path?

   • Path B, because default values are always preferred
   • Whichever path arrived first
   • Path A, because a higher local preference value wins and it is compared before AS path length
   • Neither, since local preference only affects eBGP paths

2. Weight and local preference tie between two candidate paths. One prefix was injected with a network statement on the local router; the other was learned from a BGP peer. Which one wins at this stage of the algorithm?

   • The peer-learned path, because externally learned routes are always trusted more
   • The locally originated path, because routes the router originates itself outrank ones learned from a neighbor
   • They remain tied and both stay in the table as multipath
   • The path with the lowest router ID

3. Weight, local preference, and local origin all tie. Path A has AS_PATH "65001 65002" and Path B has AS_PATH "65003 65004 65005". Which path wins the AS path comparison?

   • Path B, since a longer path implies more redundancy
   • Path A, because BGP prefers the path with the fewest AS hops
   • Both are equally valid and BGP picks at random
   • Neither; AS path length is only a tie-breaker for MED

4. Two paths remain tied through AS path length. One carries origin code "i" (IGP), and the other carries origin code "?" (incomplete). Which is preferred?

   • The incomplete path, since it typically comes from a more direct redistribution
   • They are treated identically because origin code is informational only
   • The IGP-origin path, because origin type ranks IGP above EGP, and EGP above incomplete
   • Whichever path has the numerically lower origin code value in hexadecimal

5. Two paths from the same neighboring autonomous system tie through origin type. One advertises MED 20 and the other MED 80. Assuming default bgp always-compare-med is not needed here, which path wins?

   • The path with MED 80, because BGP treats a higher MED as more trustworthy
   • The path with MED 20, because lower MED values are preferred and comparison applies since both paths share the same next-hop AS
   • Neither, because MED is never compared unless bgp deterministic-med is off
   • MED is ignored whenever local preference is present

6. All earlier best-path attributes tie between an eBGP-learned path and an iBGP-learned path for the same prefix. Which one does the router prefer?

   • The iBGP path, since internal paths are inherently more stable
   • The eBGP path, because externally learned paths are preferred over internally learned paths at this step
   • Whichever was received most recently
   • The router refuses to install either and blackholes the prefix

7. Two iBGP paths remain tied after the eBGP-versus-iBGP step. The IGP metric to the BGP next hop of Path A is 5, and to the next hop of Path B is 40. Which path is preferred?

   • Path B, since a higher IGP metric shows a more thoroughly verified path
   • Path A, because BGP prefers the path whose next hop is reachable at the lowest IGP cost
   • Neither, since IGP metric only matters for redistributed routes
   • The tie is broken solely by router ID at this point, skipping IGP metric

8. Two eBGP paths remain tied even after the IGP-metric comparison. One eBGP session has been up and stable for several hours; the other reset a few minutes ago. What can influence the outcome at this stage?

   • BGP always prefers the more recently established session
   • With default settings, BGP can prefer the path from the older, longer-established eBGP session as an additional tie-breaker
   • Session age is never used anywhere in the best-path process
   • The newer session automatically becomes administratively down

9. After every earlier attribute ties, two remaining eBGP paths are advertised by routers with BGP router IDs 3.3.3.3 and 8.8.8.8. Which one wins the router ID tie-breaker?

   • The path from router ID 8.8.8.8, because a higher router ID always wins
   • The path from router ID 3.3.3.3, because the lower router ID is preferred at this step
   • Both remain installed as ECMP regardless of router ID
   • Router ID is never used as a tie-breaker in modern IOS-XE

10. IP connectivity between two intended BGP peers is confirmed with ping, but the BGP session never leaves Idle or Active. An access list on the path denies TCP destination port 179. What is the most likely explanation?

    • OSPF has not converged between the two routers
    • BGP cannot establish its TCP session because the ACL is blocking the well-known BGP port
    • The routers have mismatched autonomous system numbers only
    • MPLS labels are not being exchanged

11. Two eBGP routers are configured to peer using their loopback interfaces. ebgp-multihop and routes to both loopbacks are in place, but neither side applies neighbor <ip> update-source Loopback0. What symptom results?

    • The session forms fine, but MED is ignored
    • Each router sources its TCP connection from its outgoing physical interface instead of the loopback, so the peer rejects the unexpected source address and the session never establishes
    • AS path prepending stops working
    • The route reflector cluster ID becomes invalid

12. Two eBGP routers peer via loopback addresses that are more than one Layer 3 hop apart, and the session fails with the TTL expiring in transit. What is required to fix this?

    • Configure neighbor <ip> ebgp-multihop <ttl> to raise the TTL and permit the extra hops
    • Configure next-hop-self on both routers
    • Lower the eBGP AS path length limit
    • Disable PHP on the intermediate LSR

13. An iBGP-learned prefix shows in the BGP table but is marked inaccessible because its next hop cannot be resolved through the IGP. Which fix is most appropriate at the edge router that originally learned the route from an eBGP peer?

    • Apply neighbor <ip> next-hop-self so the edge router advertises itself as the next hop to its iBGP peers
    • Increase the MED sent to the iBGP peers
    • Set local preference to 0 on the affected prefix
    • Disable synchronization globally

14. Which command gives the fastest condensed view of whether IPv4 unicast BGP neighbors are established and how many prefixes each is sending or receiving?

    • show ip route bgp
    • show bgp ipv4 unicast summary
    • show ip bgp neighbors detail
    • show ip protocols

15. A route arrives tagged with the well-known BGP community no-export. What is the correct handling of that route by the receiving AS?

    • It must be advertised to every eBGP peer regardless of policy
    • It must not be advertised outside the receiving autonomous system, including not to eBGP peers (unless the AS is a confederation, where sub-AS boundaries are treated specially)
    • It must be dropped immediately and never installed in the RIB
    • It forces the receiving router to set local preference to 0

16. A route is marked with the well-known community no-advertise. What behavior does this enforce?

    • The route can be sent to eBGP peers but not iBGP peers
    • The route must not be advertised to any BGP peer at all, internal or external
    • The route is only withheld from route reflector clients
    • The route is advertised normally but with MED set to the maximum value

17. A route-map sequence needs to match routes carrying a specific BGP community value before applying a set action. Which supporting object should the route map reference?

    • A standard access list matched with match ip address
    • A community list referenced with match community
    • A prefix list matched with match ip address prefix-list
    • An AS path access list matched with match as-path

18. A peer starts advertising far more prefixes than the configured neighbor <ip> maximum-prefix <limit> threshold. What is the router enforcing when it logs a prefix-count warning or tears the session down?

    • The BGP maximum-prefix limit, a protection against accidental or malicious route floods from that peer
    • The eBGP multihop TTL
    • An implicit AS path access list
    • The route reflector cluster boundary

19. A router is configured with aggregate-address 10.10.0.0 255.255.0.0 summary-only. What happens to the more-specific component routes that fall inside that range?

    • They continue to be advertised alongside the aggregate
    • They are suppressed from advertisement; only the aggregate route is sent to peers
    • They are advertised only to iBGP peers, never eBGP
    • They are withdrawn from the local RIB entirely

20. A large enterprise with 40 iBGP routers wants to avoid configuring a full iBGP mesh. Which design directly reduces that iBGP peering burden?

    • Enabling BGP synchronization on every router
    • Deploying route reflectors so clients peer only with the reflector instead of every other iBGP router
    • Lowering the local preference on all internal routes
    • Switching every internal session from iBGP to eBGP

21. Router R1 is a plain iBGP speaker (not a route reflector) with two iBGP neighbors, R2 and R3. R1 learns a prefix from R2 over iBGP. Why does R1 not forward that route to R3?

    • R3 has a lower router ID than R2
    • The iBGP split-horizon rule blocks readvertising an iBGP-learned route to another iBGP peer, to prevent routing loops without full-mesh knowledge
    • R1 has no FIB entry for the prefix
    • AS path prepending strips the route before it reaches R3

22. A route reflector environment has redundant reflectors serving the same client cluster. Which attributes let a reflector detect and drop a route that has looped back to it?

    • Weight and local preference
    • MED and origin type
    • Originator ID and cluster list
    • Next hop and AS_PATH length

23. A BGP update carries the atomic aggregate attribute. What does that attribute communicate to downstream routers?

    • The route was learned via eBGP multihop
    • Some more-specific path information was lost when the aggregate route was formed, so downstream routers should not assume all contributing paths shared identical attributes
    • The route is protected by BFD
    • The route must be treated as a default route

24. BGP multipath is enabled and two equally good eBGP paths to the same prefix are now both installed in the RIB for load sharing. What remains true about the best-path selection process itself?

    • BGP no longer runs the best-path algorithm at all once multipath is enabled
    • BGP still runs its normal best-path comparison and designates one path as best; multipath then also installs the other equally ranked eligible path or paths alongside it
    • Only the newest path is kept and the older one is withdrawn
    • Multipath disables MED comparison entirely for all prefixes