SOA C90.03 問題集

質問 1：
Cloud Service Consumer A accesses Cloud Service A (1) that resides in Cloud X.
a private cloud owned by the same organization acting as Cloud Consumer A.
Cloud Service A
processes the message from Cloud Service Consumer A and then sends back a response
with the requested data (2). Next, Cloud Service Consumer A sends a message containing
some of this data to Cloud Service B (3), which resides in public Cloud Y that is owned by
Cloud Provider Y.
After processing the message. Cloud Service B sends back a response
with additional data to Cloud Service Consumer A (4). Finally, Cloud Service Consumer A
writes the data it collected from Cloud Services A and B to Database A (5).
Recently, Cloud Service Consumer A has been required to access Cloud Services A and B
at a significantly higher rate, sometimes over 1,000 times within a given workday. This
increased usage has not affected Cloud Service B' s performance. Cloud Service A,
however, has been generating runtime exceptions, and responses to Cloud Service
Consumer A have become increasingly slow and unreliable. It is determined that this
decline in performance is due to infrastructure limitations within private Cloud X' s
environment. Instead of investing in new infrastructure for Cloud X, it is decided to explore
the feasibility of moving Cloud Service A to Cloud Y instead.

Which of the following statements describe valid financial considerations that can be taken
into account for assessing the feasibility of this move?
A. Because, in this scenario, the cost of capital is comprised of the up-front costs added to
the on-going costs, the cost of capital required to move Cloud Service A to Cloud Y will be
higher than upgrading Cloud X to accommodate Cloud Service A' s increased usage.
B. By moving Cloud Service A to Cloud Y, there may be a decrease in operational
governance control over the Cloud Service A implementation. This can increase locked-in
costs because Cloud Consumer A may be forced to form dependencies upon proprietary
tools used to configure and maintain the Cloud Service A implementation.
C. If the existing infrastructure that currently resides in private Cloud X was purchased
specifically in support of Cloud Service A, then there may be a financial loss resulting from
moving Cloud Service A out of Cloud X.
This can be considered sunk costs that need to be
evaluated.
D. Moving Cloud Service A to Cloud Y will result in the need for Cloud Service A to
undergo integration testing to determine how well it can function within Cloud Y and what
changes may need to be made to Cloud Service A in order for it to behave as expected.
The integration testing and the subsequent changes required for Cloud Service A to
function correctly within Cloud Y will incur integration costs that need to be budgeted for.
正解：B,C,D

質問 2：
The cloud service owner of Cloud Service A is evaluating Clouds X, Y and Z to determine
which cloud environment can offer the greatest level of reliability. All three clouds are
geographically dispersed across three separate time zones. As a result, each cloud
experiences usage peaks at different times. Based on the metrics provided, the greater the
usage of a cloud, the lower its reliability. When the cloud service owner complains to Cloud
Provider A (the owner of all three clouds) that none of the clouds provide an adequate level
of reliability, Cloud Provider A suggests a solution that increases resiliency.

Which of the following statements accurately describes a solution that can be used to fulfill
the resiliency requirements of Cloud Service A?
A. A cloud balancing solution is established, whereby a resource replication mechanism is
implemented on each cloud. This allows Cloud Service A to be automatically replicated
across cloud environments, thereby enabling each implementation of Cloud Service A to
take the place of another, whenever failure conditions occur.
B. A cloud balancing solution is established, whereby an automated scaling listener
mechanism is implemented on each cloud in such a way that every cloud can automatically
scale out to another cloud. As a result, if reliability problems occur on any one cloud, the
subsequent requests will be scaled out to another cloud in a manner that is transparent to
cloud service consumers.
C. A failover system mechanism is implemented on Cloud X, which acts as the primary
point of contact for cloud service consumers. Upon failure conditions occurring, the Cloud
Service A implementation on Cloud X automatically hands over control of current and
future message requests from cloud service consumers to Cloud Y.
Cloud Y retains control of cloud service consumer communication until the next failure condition occurs, at which
point it hands over control to Cloud Z.
Finally, if a failure condition occurs in Cloud Z.control is handed back to Cloud X.
D. Redundant implementations of Cloud Service A are deployed in all three clouds. The
failover system mechanism and a special type of automated scaling listener mechanism
are implemented to establish a system whereby one redundant Cloud Service A
implementation will automatically take over from another.
正解：D