ExamGecko
Question list
Search
Search

List of questions

Search

Related questions

Question 5 - JN0-683 discussion

Report
Export

Exhibit.

You are troubleshooting an IP fabric (or your data center. You notice that your traffic is not being load balanced to your spine devices from your leaf devices. Referring to the configuration shown in the exhibit, what must be configured to solve this issue?

A.
The load-balance policy must be applied to the forwarding table under the routing-options hierarchy.
Answers
A.
The load-balance policy must be applied to the forwarding table under the routing-options hierarchy.
B.
The multipast multiple -as configuration must be configured for each peer in the BGP spine group.
Answers
B.
The multipast multiple -as configuration must be configured for each peer in the BGP spine group.
C.
The load-balance policy must be applied as an export policy to your BGP
Answers
C.
The load-balance policy must be applied as an export policy to your BGP
D.
The load-balance policy must have a from statement that matches on protocol bgp.
Answers
D.
The load-balance policy must have a from statement that matches on protocol bgp.
Suggested answer: B

Explanation:

IP Fabric Load Balancing:

In the provided configuration, traffic is not being load-balanced to the spine devices. The issue likely relates to how BGP routes are being selected and whether Equal-Cost Multi-Path (ECMP) is functioning correctly.

Multipath Multiple-AS:

Option B: The multipath multiple-as configuration is essential when using BGP in an IP fabric where devices belong to different Autonomous Systems (AS). This setting allows BGP to consider multiple paths (even across different AS numbers) as equal cost, enabling ECMP and proper load balancing across spine devices.

Conclusion:

Option B: Correct---The multipath multiple-as configuration is necessary for achieving ECMP and effective load balancing in a multi-AS BGP environment.

asked 18/09/2024
Hans Walter Katzengruber
27 questions
User
Your answer:
0 comments
Sorted by

Leave a comment first