Practice Question R & S

[CiscoWizard]

I came across this practice question and the answer, which I thought was obvious, was actually incorrect according to the site.

What is the path for traffic sent from one user workstation to another workstation on a separate switch in a three-tier architectural model?
A. Access - Core - Distribution - Access
B. Access - Distribution - Distribution - Access
C. Access - Core - Access
D. Access - Distribution -  Core - Distribution - Access

I chose B as the answer. The correct answer is listed as D. It says the other workstation is on another switch, not another network. Why would you need to bring the Core layer into it? And since Distribution layer switches can also be layer 3 switches, you would definitely not need the Core layer.

What does everyone think?

[Otanx]

With the options presented I would go with D. Lets follow the normal test method. Both A and C are obviously wrong. Access does not connect to Core in the three tier model. I see B as wrong because you wouldn't have a Distribution switch connected directly to another Distribution switch. Just like you wouldn't have two Access layer switches connected together. So if we need to go between two different Distribution switches then we need to use the Core, and that is option D.

-Otanx

[CiscoWizard]

I see the logic in that but I've seen plenty of networks that create the redundancy at the Distribution Layer to avoid using the Core layer for unnecessary traffic. We have over 50,000 employees in my company. The networks are broken down into multiple locations. In the building where I am located we have two main Core devices that head out to the Internet (and to other company locations), but two Distribution layer devices on each floor which act as a backup for each other. We try to keep traffic between two people on the same floor and other floors out of the Core portion of the network. I know others who used the same setup. I feel like the answer being B or D can be relative. At least we do agree both A and C are wrong. 

[icecream-guy]

The question is vague, it could be B or D depending on architecture.

[Otanx]

I see the logic in that but I've seen plenty of networks that create the redundancy at the Distribution Layer to avoid using the Core layer for unnecessary traffic. We have over 50,000 employees in my company. The networks are broken down into multiple locations. In the building where I am located we have two main Core devices that head out to the Internet (and to other company locations), but two Distribution layer devices on each floor which act as a backup for each other. We try to keep traffic between two people on the same floor and other floors out of the Core portion of the network. I know others who used the same setup. I feel like the answer being B or D can be relative. At least we do agree both A and C are wrong. 

Yep, but those cross connects should be in a blocking state assuming the cores are the roots if you have a layer 2 topology. If you are thinking full layer 3 then you could make the argument that B could be right, but D would be right in more cases. So the most right is D. This is the fun of IT exams. There can be more than one right answer, and you need to pick the 'most right' based on what the exam writer thinks.

-Otanx

[CiscoWizard]

What I do like about these questions that I feel can have more than one correct answer (and this isn't the first), it implores me to research the answers further and I come away from it learning a lot more. The way you describe it makes sense of course. If the switches were using layer 2 then the redundant link would be blocking. If they're using layer 3, then it's possible to reach the other device through the Distribution layer switch. They don't give you a lot of details as far as the users' networks.

[deanwebb]

D is the correct answer because it has all three tiers and, therefore, more Cisco gear purchased, including a tasty Nexus 9K at the core.

When there are two good choices from a technical view, the correct answer is the one Cisco marketing and sales are pushing.

[CiscoWizard]

In that case there should be a choice E.

E. Access - Distribution - Core - NGFW - NGFW (backup) - Core - Core (backup) - Distrocore Sub-Layer - Distribution - Distroaccess Sub-Layer - Access

[deanwebb]

In that case there should be a choice E.

E. Access - Distribution - Core - NGFW - NGFW (backup) - Core - Core (backup) - Distrocore Sub-Layer - Distribution - Distroaccess Sub-Layer - Access

They'd rather leave the NGFW out, that might wind up pushing Palo gear.

[CiscoWizard]

[deanwebb]

Win.

You're gonna like it here!