Architecture Recommendation
Given a fictional enterprise scenario, produce and defend a recommendation — upstream, a named enterprise distribution, another traditional NOS, or a mixed model — justified by risk, staffing, support, features, and total operational cost including labor. The deliverable is a written recommendation backed by the Lesson 5.4 scorecard.
Non-disruptive. Safe to run in the virtual lab.
This is a written exercise; there is no device interaction and no risk to any environment.
From scenario to defended recommendation
Objective
- • Characterize a fictional organization's capacity, risk tolerance, scale, and constraints.
- • Apply the Lesson 5.4 scorecard and produce a defensible recommendation.
- • Ground the justification in labor and operational ownership, not license price.
Starting state
- • A fictional scenario in hand (use the one below or one provided by your instructor).
- • A blank copy of the Lesson 5.4 scorecard.
Prerequisites: Lessons 5.1–5.6 complete.
Tasks
- 1
Read and characterize the scenario
Example scenario: 'Meridian Freight operates two regional data centers, runs a 6-person network team with no software-build experience, is subject to industry regulation requiring documented security advisories and audit trails, currently runs Dell OS10, and wants to modernize its leaf-spine fabric with EVPN/VXLAN over the next 18 months on a predictable budget.'
Extract: engineering capacity, risk tolerance, regulatory constraints, scale, existing vendor relationship, and timeline.
Expected:- A short written profile of the organization's capacity, constraints, and priorities.
What happened internally: The recommendation must follow from the organization's actual capacity — this task forces you to name it before choosing.
- 2
Fill the scorecard for each candidate
Score upstream, at least one enterprise distribution, and 'keep/extend the traditional NOS' against the Lesson 5.4 rows, including operational-staffing and TCO-with-labor.
Expected:- A completed scorecard per candidate with go/no-go rows marked and TCO including labor.
What happened internally: Forcing the staffing and TCO-with-labor rows is what prevents a naive 'upstream is free' conclusion.
- 3
Write the recommendation
State the chosen model (upstream / named enterprise distribution / traditional NOS / mixed) in one sentence, then justify it from the scorecard.
Expected:- A clear recommendation justified by risk, staffing, support, features, and total operational cost including labor.
What happened internally: For the example scenario, a small team with no build experience under regulation typically points to an enterprise distribution or staying on the traditional NOS — but the justification, not the label, is what is assessed.
- 4
Name the accepted risks and a review trigger
List the risks the recommendation accepts, their mitigations, and when the decision should be revisited.
Expected:- A short risk-and-mitigation list plus a defined review trigger (e.g. staffing change, scale threshold, vendor release).
What happened internally: Every model carries residual risk; naming it and setting a review trigger is what makes the recommendation professional rather than dogmatic.
Troubleshooting branches
If: The recommendation feels forced toward SONiC
Then: 'Keep the traditional NOS' and 'mixed model' are valid outcomes; if the scorecard points there, say so.
If: Justification leans on license price
Then: Re-center on staffing and operational ownership; license is one row, not the argument.
- None — this is a written deliverable.
Lab evidence checklist
- ☐ Organization profile written (capacity, constraints, priorities).
- ☐ Scorecard completed for each candidate, including staffing and TCO-with-labor.
- ☐ One-sentence recommendation plus scorecard-based justification.
- ☐ Accepted risks, mitigations, and a review trigger documented.
Reflection
- If the organization's team doubled in size and gained build experience, would your recommendation change — and what does that reveal about the decisive variable?
- Where did license price try to dominate your reasoning, and how did the staffing/ownership rows correct it?