a.

Non-negotiable Critical Constraints for Cloud Strategy Design

In designing the cloud strategy for such a medium-sized enterprise, the cloud
architect must consider the following non-negotiable constraints:
1. Legal and Regulatory Compliance: The firm has to ensure that, based on its
industry and geography, it complies with regulations like data residency laws,
GDPR, HIPAA, or other industry-specific regulations. Ensuring that the cloud
environment aligns with these legal frameworks is not up for negotiation.
2. Data Privacy and Security: Business and customer data are sensitive; hence,
they need to be protected. An architect has a prime responsibility to ensure that
security cloud provider encryption, access controls, and incident response offer
protection against unauthorized access or data breaches.
3. Geography and Latency: Places of data centers must be carefully chosen
because native laws may have an impact on where data is kept and processed.
Geographical limitations may also involve latency and performance requirements
depending upon the proximity of users to data centers.
4. Business Continuity and Disaster Recovery: Architecture shall provide for a
credible disaster recovery solution to ensure least or minimum downtime.
Systems shall have redundancy, backup, and failover capabilities to protect
against service interruptions.
5. Cost Control: Economic constraints need to be put into consideration,
especially regarding capital and operational expenditure. Moving to CAPEX to
OPEX models with cost efficiency is important.
b. Risks and Economic Challenges of a Multi-cloud Strategy
A multi-cloud approach can offer flexibility and resilience, yet it also introduces
some particular risks and economic challenges:
1. Complexity: Multiple cloud environments, including AWS, Azure, and Google
Cloud, present operational complexity. The APIs are different, as are the
management tools and pricing models. This might affect IT staff with extra
overhead and perhaps inefficiencies that could arise.
2. Integration Challenges: The integration of different cloud platforms requires
compatibility between services, APIs, and architectures. It can lead to heightened
time and cost needed for smooth integration, bringing possible issues in
interoperability.
3. Security Risks: Security management across more than one platform leads to
anomalies. The providers may support different security standards, rendering the
implementation of uniform security policies challenging.
4. Cost Overruns: While multi-cloud prevents vendor lock-in, it will increase the
cost of using extra management tools and resources besides retaining expertise
in multiple platforms. Differences in pricing across providers may complicate
budget forecasting.
Risk Addressing:
• Centralized Management: CMP would be implemented to monitor, within one
location, control, and automate activities across multi-clouds.
• Security Framework: Deploy a robust security governance framework
consistently on all platforms to reduce security risks.
• Workload Optimization: Leverage each cloud provider’s strengths. For example,
use one provider for compute-intensive workloads and another for data storage,
optimizing costs and performance.
• Cross-Platform Knowledge: Invest in internal knowledge across platforms, or
leverage managed services to take care of the intricacies of a multicloud
environment.
c. Gartner's Six Steps for Cloud Strategy Planning
The cloud architect should focus on Gartner's six steps in developing an
appropriate strategy for the enterprise, considering all company goals with
respect to cost, agility, and risk:
1. Velocity and Business Value: First, the architect needs to define how the cloud
strategy shall deliver business value, with a focus on driving the agility of the
company in response to market changes and demand from customers. Speed is
essential regarding the deployment of new services as well as infrastructure to
meet customer demand.
2. Ensure Resilience: It's about cloud environment resilience and reliability to
make sure risks are at their lowest. Architecture must be fault-tolerant and
support automatic backups, disaster recovery, and redundancy.
3. For any business on multi-cloud, there should be a primary provider, even as
they work with multiple providers. It is very important to have a primary cloud
provider that handles core services, supplemented by secondary providers when
required for ancillary uses like disaster recovery or specialized workload.
Sometimes, this allows for easier and less complex management.
4. Optimize for Cost: This shall involve the architect focusing on optimizing the
cost, ensuring the enterprise applies auto-scaling to avoid paying for resources
when not used, and optimizing resources to avoid platform costs.
5. Ensure Agility: Agility is crucial for the enterprise to adapt to market changes
or customer needs. The architect must ensure the cloud infrastructure is
designed for rapid scaling and flexibility, enabling the company to quickly
respond to new opportunities or challenges.
6. Manage Risk and Compliance: Finally, the architect needs to ensure that risk
and compliance requirements are in place to monitor at all times and ensure that
the company is adhering to regulatory requirements, standards of data security,
and protocols of internal risk management.

The company wants to move into the cloud and understand from that shift in
direction how it's going to affect their return on investment and day-to-day
operational expenses. Basically, what they will want to see is how to minimize
costs and still have fast, efficient systems, which any e-commerce platform
needs since a slow website would turn away customers. The cloud architect
comes into play to help them balance between CAPEX and OPEX during this
transition.
a. Key Drivers of Cloud Computing ROI
1. Use-Based Pricing: Cloud services provide a pay-as-used basis, reducing initial
capital investments. Businesses only pay for resources utilized, which could
reduce overall costs drastically compared to traditional infrastructure with fixed
capacity.
2. Scalability and Flexibility: Cloud resources can be scaled up or down
depending upon demand. This elasticity decreases the need to invest in
expensive capital hardware upfront, thereby minimizing underutilized resources
to optimize returns.
3. Operational Efficiency: By automating many tasks related to infrastructure
management, such as updates, scaling, and backups, cloud adoption frees up
technical teams to focus on innovation, which can further drive productivity and
profitability.
4. Capital Cost Avoidance: Adoption into the cloud frees the business from
making huge capital expenses in hardware, storage, and physical infrastructure,
and all these are shifted to operational expenses. Thus, this decreases the initial
financial burden and involves better financial flexibility.
b. Decrease Capital Investments while assuring high performance and scalability.
The cloud architect can give several recommendations to help the business
reduce capital investments without compromising high performance and
scalability:
1. Adopt a Hybrid Cloud Environment: The organization shall maintain critical
systems on-premises while moving less critical applications to the cloud. In this
case, the migration can be gradual with minimal CAPEX but scalable in the cloud
for performance.
2. Leverage the Autoscaling Features: The cloud autoscaling automatically
enables the system to scale up during high demand and scale down during low
traffic. This way, it ensures that the business maintains performance without
over-provisioning and reduces unnecessary spending on infrastructure.
3. Choose the Right Cloud Service Model-IaaS, PaaS, SaaS: Based on the business
needs, it is he who should suggest the selection of a service model that could
bring economy. For example, Platform as a Service-PaaS reduces investment in
infrastructures for the development and deploys applications faster.
4. Containerization and Serverless Architecture: The use of containerized
workloads, or a serverless architecture like AWS Lambda, Google Cloud
Functions, etc., will make application executions possible in an isolated
environment efficiently by the business. These types of solutions also give on-
demand performance without the need to provision and maintain entire servers.
5. Optimization of Workload and Resources: Further workload optimization by the
architect involves the identification of applications that can be hosted on
cheaper cloud environments or less powerful infrastructure during periods of low
usage to minimize costs while keeping critical workloads high performing.
c. Financial Planning of CAPEX and OPEX Trade-offs
The following must be considered to balance CAPEX by the business:
1. Progressive Adoption of Cloud: The cloud architect will explain a strategy that
strikes a balance between CAPEX and OPEX, leaving mission-critical systems on-
premises-taking the CAPEX-while moving less critical workloads, such as
customer service or test environments, out to the cloud. In so doing, immediate
needs for CAPEX are reduced while flexibility and scalability in OPEX increase.
The company will thus be able to shift workloads to cloud in due course at an
easier transition on financial levels.
2. Leveraging Reserved Instances for Predictable OPEX: The business should
pursue reserved instances/long-term cloud contracts offering substantial
discounts on cloud services. This approach would lock in lower rates on services
that a business knows it will use, therefore providing greater predictability in
monthly expenses and lowering overall OPEX.
3. Monitoring Operational Costs: Cloud environments introduce variable costs
based on usage. Periodically, the business has to go through the cloud bills and
their usage data periodically for the estimation of operational expenses with high
accuracy. Also, it includes making a financial plan keeping in mind the predicted
peaks of resource demand, such as sales events or seasonal traffic, and enabling
cost-savings like autoscaling and resource optimization.
To balance CAPEX and OPEX, the business needs to ensure that proper
governance and monitoring mechanisms are put in place to avoid overspending
in a cloud environment.

ABC Corp is a global organization moving to cloud-based solutions but faces

problems in IT governance and change management. The main goal here will be
to align the IT strategy with overall business objectives and prepare the
employees effectively for this major transition.
a. Potential Benefits of a Strong Governance Framework
1. Decreasing Risks: A well-defined governance framework aids the company in
recognizing possible perils quite early and coming up with straightforward
policies for mitigating them. This involves dealing with security risks, compliance
regulations, such as GDPR, industry standards, and ensuring data integrity.
Proper governance ensures that operational risks such as service downtime or
data breaches are minimal.
2. Driving Economic Growth: Governance aligns IT strategy with business
objectives, ensuring the clouds' investments directly contribute to economic
growth. By prioritizing and optimizing IT investments, the company reduces
inefficiencies and better allocates resources to support strategic initiatives that
drive growth. This, in turn, maximizes the return on cloud investments.
3. Retaining Investor Confidence: Good governance indeed ensures that
investors are confident in the transition into the cloud, hence ensuring there are
no financial mismanagements or operational disruptions. Transparency in
decision-making and reporting will ensure that stakeholders understand how the
company is leveraging cloud solutions to realize long-term goals.
b. Problems in Managing Cultural Changes and Change Management's Role
1. Resistance to Change: Employees who are accustomed to working with on-
premises systems may resist moving to cloud-based workflows due to fear of the
unknown or uncertainty about their roles in the new system. This resistance
could hinder the adoption of cloud technologies.
2. Inadequate Cloud Knowledge and Skills: Most employees lack adequate skills
for working in cloud environments. The gap in skills makes the transition of
operations to the cloud more frictional, thus slowing the process down.
3. Fear of Job Displacement: There is the probability that staff may fear
automation in cloud adoption could make their roles redundant, hence causing
job insecurity that may create resistance and anxiety during the transition.
Relentless Change Management can address these challenges through:
• Communication: Regular and transparent communication allows staff to
understand the rationale behind the change and how this fits with the company's
goals. It may also reduce anxiety on job security by demonstrating what their
role will be within the new system.
• Training and Upskilling: Imparting focused training programs would help
employees in cloud-related skills. It will not only help put them at ease regarding
technical requirements but also provide security in retention of their jobs.
• Accountability Culture: Clearly defining expectations and creating an
accountable atmosphere from all levels of the company will encourage
employees to own the transition. Recognition and reward may be very
instrumental in revolutionizing the mindsets of people in the company.
c. Practices for Equipping Employees with Cloud Adoption Readiness
1. Early and Consistent Communication: The ability to communicate early and
consistently about the changes is important in helping employees understand
things in context-that cloud transition is required to be done for meeting certain
business ends. In fact, the openness of this builds confidence and removes
uncertainties that help reduce resistance to change.
2. Extensive Training Programs: Indeed, developing training programs will be
crucial in preparing employees, with a focus on those very important cloud-
specific skills. The training shall consider both the technical features of the cloud
systems and the new workflows that shall be adopted. Moreover, the capacity for
hands-on learning-such as cloud certification or workshops-may help employees
feel more confident in their new roles.
3. Creating a Collaborative Atmosphere: Such cross-functional collaboration will
make sure that employees across the teams jointly solve the challenges relating
to this transition. This creates a unified approach and minimalizes the chances of
silos being formed inside and outside of cloud-savvy teams.
4. Gradual Implementation and Pilot Programs: ABC Corp. should not try to shift
all the systems to the cloud at once but use gradual and phased implementation.
By beginning with less critical systems or a pilot program, employees can
gradually get adapted piece by piece without feeling the full brunt of transition
pressure.

AWS Services: Cloud Computing

a. SaaS, PaaS, and IaaS Services Offered by AWS
1. Software as a Service: AWS provides a vast array of SaaS applications,
whereby the customer utilizes software over the internet without having to
manage the underlying basic infrastructure.
o Example: Amazon Chime is an AWS Communication Service that lets the users
chat, video conference, and screen share within one fully encrypted and secure
environment. Amazon Chime is also a completely managed SaaS in which
everything, from scaling to security, is handled by AWS.
2. Platform as a Service: AWS also incorporates PaaS solutions that provide
developers with a platform on which they can develop, deploy, and manage
applications without being involved in the management of infrastructural mess.
o Example: AWS Lambda is a PaaS that allows developers to run code without
any requirement to provision or manage servers. It runs code on occasions in
response to events automatically scales applications and only charges the user
for the compute time used. This serverless approach will hence turn perfect for
event-driven applications that involved real-time data processing or analytics.
3. IaaS: With AWS, IaaS offerings offer basic building blocks, such as computing
power, storage, and networking resources on-demand.
oExample: AWS EC2 is an Elastic Compute Cloud-a popular IaaS service where
users rent virtual servers to run their applications. It provides scalable computing
capacity, whereby businesses can scale up or down depending on the workload.
b. AWS Deployment Models
AWS provides different deployment models for cloud computing, which fulfill
various business needs:
1. Public Cloud: AWS provides a public cloud, which delivers services over the
internet and is shared between multiple organizations. This model is the most
common for businesses seeking cost-effective, scalable solutions without the
need for managing infrastructure.
o Example: Amazon S3 (Simple Storage Service) A Public cloud service that
provides object storage with scalable features to any AWS customer. It has very
high durability that enables the organization to store large data in various
regions.
2. Private Cloud: AWS also supports private cloud environments where resources
are used by a single organization exclusively to provide control and security for
sensitive workloads.
o Example: AWS Outpost allows an organization to run AWS services on-premise
while reaping the advantages of the AWS infrastructure. This ensures a hybrid
environment where businesses can deploy resources privately but still integrate
with AWS services.
3. Hybrid Cloud: AWS does support hybrid clouds through which an organization
can combine on-premise infrastructure with cloud resources to achieve flexibility
and scalability.
o Example: AWS Direct Connect is a hybrid deployment model. This model
facilitates enterprise businesses to establish a dedicated network connection
between their on-premise data centers and AWS. In this, lower latency and more
secure connections are provided compared to the standard internet-based
access.

Public Cloud and Private Cloud: Adoption of Cloud and Workload Preference ABC
Inc. is looking to adopt either a public or private cloud to retire its in-house
infrastructure. To help ABC come to a reasonable decision, let's delve into the
different Cloud adoption and workload preference levels presented by Public and
Private Clouds:
Public Cloud:
Public cloud cost-effectiveness can be unlocked via pay-as-you usage with
reduced upfront investments, which best fits scenarios where the workload is
variable in nature and the budgets are smaller. It offers great scalability and
flexibility for fitting resources to demand, which helps in seasonal or fluctuating
workloads. Public cloud services are global in reach and hence can provide
access to data centers around the world. Of course, lower latency and regional
regulations are complied with, while security is robust. However, multi-tenancy
could raise several issues regarding infrastructural sharing.
Private Cloud:
With private cloud environments, companies also have more control over how
much they can customize to suit particular needs in infrastructure and security
policy. This fits for organizations that have strict regulatory demands and those
dealing in sensitive data. Private clouds will be predictably consistent in
performance, with resources not shared with other companies, which will make it
fit for workloads that demand high availability.
On the other hand, private cloud solutions are costlier in that they include high
initial investments and consequent maintenance costs, which usually are
justified in instances where the entity requires more control, security, and
personalization.
Workload Preferences:
Workloads best-suited to the public cloud are dynamic, scalable applications,
such as web hosting, data analyses, and development/test environments. This is
because the latter take advantage of cost efficiency and flexibility. Sensitive
workloads subject to strict compliance would better fit in with the private cloud,
such as healthcare systems and financial services, whereby security in data and
consistent performance are more crucial.
Recommendation for ABC Inc.:
• In case ABC Inc. prioritizes cost savings, flexibility, and scalability, it would be
best to have its workloads insensitive or non-compliant on public clouds. These
options will help ABC Inc. scale up and down the resources with demand while
economically efficient.
• However, the private cloud would be more suitable in case ABC Inc. is handling
sensitive data or operates in a highly regulated industry where full control of the
infrastructure needs to be at its discretion, notwithstanding the overall higher
costs. • Hybrid cloud model: In this case, ABC Inc. can also consider a hybrid
cloud model that combines the benefits of both public and private clouds to allow
the company to optimize costs while keeping critical workloads sensitive to their
control.

Migration Costs and Partitioning a. Application characteristics and their impact on

overall cost (4 pt.)
Tak, Sivasubramaniam and Urgaonkar's paper identifies a number of
characteristics that impact cloud migration cost. The first is workload variability.
Applications whose workloads are unpredictable or time-varying favor the cloud
the most because of its capability to scale the resources up or down according to
demand. This elasticity prevents over-provisioning that is commonly practiced
with on-premises systems.
On the other hand, this advantage does not occur with workloads in which the
demands are relatively stable, since they could not fully use the dynamic scaling
of the cloud.
Another attribute is resource utilization: for example, applications that are
computationally intensive or demand high CPU, memory, or disk I/O usually
involve higher costs when running in the cloud environment because of the need
to utilize more powerful and hence more expensive cloud instances. Also,
significant data transfer or I/O-intensive operations for an application may imply
additional costs, because most of the cloud pricing models charge for the data
movement between services or regions. This makes applications with heavy data
transfer requirements more expensive to operate in the cloud than their lighter
data counterparts.
Architectural design of an application is another factor influential in migration
costs. In cloud environments, distributed applications are a natural fit, wherein
application infrastructure is used to run efficiently on the cloud. On the other
hand, monolithic applications may require re-architecture to support clouds and
thus include higher costs during migration. b. Horizontal and Vertical Partitioning
(6 pts)
Horizontal partitioning, or sharding, refers to the splitting of a database into
smaller, more manageable pieces, using rows. Each partition, or "shard,"
contains a subset of the total amount of rows within an application, usually
determined by something like user ID or geographic region. This approach
facilitates the distribution of data over numerous servers, which helps balance
the load and improve performance, especially for applications with high
transaction volumes. For instance, a worldwide customer database might be
divided by region, with each region's data stored on a separate server so that no
one server would become overwhelmed by the traffic. Such a technique
improves both performance and availability by distributing queries and workload
effectively. On the other hand, vertical partitioning separates the database table
by columns instead of rows. In this technique, the columns are separated into
distinct tables, combining frequently accessed columns for faster retrieval of
data. In simpler terms, vertical partitioning is applicable in situations when not all
columns in a table are equally accessed. By storing these more frequently
accessed columns separately, it minimizes the amount of data being read during
a query and hence results in quick query response times. For example, an
application where user IDs and names are queried very frequently but additional
profile information isn't would place user IDs and names in one table, and less-
used data in another. The idea here is that both horizontal and vertical
partitioning provide various ways of performance optimization for applications
according to their data access patterns, leading to greater scalability and
efficiency at the same time-perhaps at lower costs in a cloud environment.