You are the technical designer for a vSphere platform transformation project. After conducting
SME interviews and using various platform information-gathering methods, you have created a
high-level design document.
This document specifies the following:
Requirements:
R1. The solution must not have a single point of failure.
R2. Production applications must not have an outage of more than 10 seconds.
R3. Data must be based in the UK.
R4. There is a 7-year retention policy for contracts.
R5. Applications should support existing and developing workloads for the next 3 years’ growth.
Spec of servers:
Web 1 vCPU, 2 GB RAM, 100 GB storage

App 1 vCPU, 4 GB RAM, 100 Gb storage
DB 2 vCPU, 16 GB RAM, 750 Gb storage
At a late stage in the software development life cycle of a production application developed inhouse, an unfortunate issue was identified when the application was deployed to the production
vSphere cluster. The production core stacked switch has capacity issues, and this is having a
serious impact on all applications for which the switch is providing network services. Within the
test system, the application works as intended in the single test VLAN and with a single-host
configuration.
Which of the following could be tried to help in this situation, from a vSphere perspective?
(Choose two)

A.
Redevelop the application for a virtual platform

B.
Place the application into a single-vApp network

C.
Add DRS rules to keep network traffic within the same host, where possible

D.
Configure network I/O control

Explanation:

If you limit the application traffic to a specific dedicated network (that is, a separate VLAN)
andusing enforced DRS affinity rules, the application traffic will not traverse the ESXi host’s
physical networkinterfaces. This will ensure that the impact of the application is minimized, while
also ensuring that theapplication itself is not limited.
The application servers already have vCPU settings. This suggests that the system has already
beingvirtualized in both types of environments. Network I/O control could be useful in the event of
contention, but therole of a designer would be to plan to prevent contention where possible; other
options would be morebeneficial. In addition, Requirement 5 specifies that the system should work
with workloads over the next 3years. Network I/O control would suggest contention very early in
the platform history.
Chapter 4, NetWork – Logical and Physical Design to allow applications to flow