The Cloud Computing Logical Architecture

Published Date: 02 Nov 2017

Cloud computing is an internet based computing where software, shared resources and information are served to devices such as computers, electricity grid. In these computing techniques, how the work is getting done is hidden from its users. It describes a new consumption, delivery and supplement model for IT services. It is kind of by-product that provide access to remote sites accessible through internet. Practically applications area of cloud computing not limited. It is latest technology in market.

Though there is no official definition and straight forward way to explain what exactly cloud computing is but it can be expressed in general as the following statement: "cloud computing is such a type of computing environment, where business owners outsource their computing needs including application software services to a third party and when they need to use the computing power or employees need to use the application resources like database, emails etc., they access the resources via Internet."

For instance, we may have a small business, where we need a few small servers for database, emails, applications etc. Normally, servers need higher computing power. On the other hand, PCs or laptop needs lower computing powers and are much cheaper than servers. Moreover, to maintain a client-server environment, we need to have a highly skilled network maintenance team. If we decide to avoid the need of purchasing servers and thus cut-off the need of keeping an operation and maintenance team, then going for clouding computing is a very cost-effective solutions. In cloud architecture, there is no need to install or maintain servers. Just by paying a fixed amount of monthly charge you can outsource your IT infrastructure into a third party IT managed service data center.

When someone talks about the cloud computing system, it is very helpful to divide this system into two sections, one is front end and other is backend. They are connected with each other via a network and mostly internet is used for fulfilling the requirement. The front side is the interface for the user and the back end is the cloud section for the whole system.

Front end

The front end of the cloud computing system comprises the client’s device (or it may be computer network) and some applications are needed for accessing the cloud computing system. All the cloud computing systems do not give the same interface to users. Web services like electronic mail programs control some existing web browsers such as Firefox, Microsoft’s internet explorer or safari. Other type of systems has some unique application which provides the network access to its clients. Front end is a technical term which refers to the interface though which a user can use some kind of services, so don’t get confused with this term.

Back end

Back end refers to the some physical peripherals. In cloud computing, back end is cloud itself which may encompasses of various computer machines, data storage systems, servers. Group of these clouds make a whole cloud computing system. Theoretically, any cloud computing system can include practically any type of computer machine program that can be imagined by a human being such as from video games to data processing, software development to entertainment. Usually, every application would have its individual dedicated server for services.

A central server is established which is used for administering the whole system. It is also used for monitoring client’s demand as well as traffic to ensure that everything of system runs without any problem. There are some set of rules, generally called as protocols which are followed by this server and it uses a very special type of software known termed as middleware. Middleware allows computers that are connected on network to make communication with each other.

If any cloud computing service provider has many customers, then there’s likely to be very high demand for huge storage space. Many companies that are service providers need hundreds of storage device with digital in nature. Cloud computing system requires minimum twice the units of storage devices, as the system keeps all the information of its clients. The cloud computing system must have a copy of all the data of its client’s. Making copy of data is called redundancy.

Clouding computing is no more considered as an emerging technology. Now, it’s a reality and this low-cost computing power is gaining popularity among businessmen, especially medium and small size and governmental organizations, as people are realizing the power of cloud environments. In short, in a cloud computing architecture, all the applications are not stored in a company’s hard disk, rather it resides in a third party computer and when a company needs to use application software and they access it via Internet.

2.1.2 Example of cloud computing:

A very simple example is Yahoo mail and Gmail both are using cloud computing. When we send or receive email, we never need any application software installed in our computer. We just need an internet connection to send our emails. But note, the operating cost of cloud computing is much cheaper than having a company’s own personal IT infrastructure and managed team. Security and privacy is the only concern in cloud environments, because all our files, emails, database are hosted in a third party servers in their premise.

2.1.3 Types of cloud computing

Considering the installation of network infrastructure, a cloud environment can be broadly categorized into three types- public cloud, private cloud and hybrid cloud.

Public cloud: this is the most popular type of cloud system and is considered as a main-stream cloud system by cloud computing experts. In public cloud system a third party data center provide both disk space and computing power for all the application software. Amazon web and Google apps is the two most popular public cloud computing service providers.

Private cloud: unlike public cloud, you need to set up your own data center and also bear all the installation & maintenance cost and have complete control of all your data. This system provides more security and privacy but it is more expensive cloud solution compared to public cloud.

2.1.4 Why cloud computing?

The main advantage of using cloud computing facility is that customers do not have to pay for infrastructure installation and maintenance cost. As a user of cloud computing you have to pay the service charges according to your usage of computing power and other networking resources. Moreover, you no more have to worry about software updates, installation, email servers, anti-viruses, backups, web servers and both physical and logical security of your data. Thus, cloud computing can help you focus more on your core business competency.

2.1.5 Problems with Cloud Computing

Though from operation and maintenance point-of-view cloud computing is a great cost-effective IT solution for business of any magnitude but it has at least two major concerns: i) security and ii) privacy. Besides these concerns cloud computing, a new technology in comparison to other existing computing solutions, has lots of scope of becoming a mature system as a reliable and cost-effective computing technology.

Since due to outsourcing, all the important data resides in a third party premise, there is always a concern about the trust-worthiness of the cloud service providers. Any security and privacy violation can be fatal, keeping this in mind many business owners are still to be convinced about the security and privacy issues of cloud computing.

Banking and Governmental data are most senstivie. Just think about a classified document of any Govt. agency getting leaked or user’s credit card information falling into the hands of cyber criminals. As soon as cloud privacy issues are more and more organized and strict rules and governance for cloud operation are in place, the more and more business will feel safe to opt for cloud computing.

Despite some serious privacy related drawbacks, cloud computing is a lucrative choice to improve productivity in any business environment, where IT is in high demand. To raise the security and privacy of cloud service providers, there need to be more co-operations between world governments so as we can develop a unified global rules and guidance for running a safe cloud computing service.

2.1.6 Cloud Computing Service Architecture

You must be familiar with Google apps. Actually, Google AppEngine is a computatation service known as Google’s business solution. Other big names in cloud computing services are: Microsoft, IBM, Amazon, HP and DELL.

A description...

Fig 6: Cloud Computing Stack

Mainly, three types of services are available from a cloud service provider (figure 6).

Infrastructure as a service: service provider bears all the cost of servers, networking equipment, storage and back-ups. You just have to pay to take the computing service. And the users build their own application software’s. Amazon EC2 is a great example of this type of service.

Platform as a service: service provider only provides platform or a stack of solutions for your users. It helps users saving investment on hardware and software. Google AppEngine and Force.com provide this type of service.

Software as a service: service provider will give your users the service of using their software, especially any type of applications software. Example-Google (GOOG), Salesforce.com (CRM), NetSuite (N)

2.1.7 Understanding cloud computing architecture models

The biggest challenge in cloud computing may be the fact that there's no standard or single architectural method. In fact, there are few definitions of the cloud computing concept that are fully accepted. Therefore, it's best to view cloud architectures as a set of approaches, each with its own examples and capabilities. A cloud computing system is a set of IT resources designed to be allocated dynamically to run applications, rather than be assigned a static set of applications as is the case in client/server computing. In a cloud computing environment, a user (via a virtual desktop, for example) requests for information from an application, and it uses the resources available in cloud to run that application.

Cloud computing models

Cloud computing is new and its potential is exciting and its applications are limitless. By guaranteeing that cloud resources link effectively with virtual desktop capabilities, enterprises can get the most out of cloud services that are available from third parties. They can also efficiently redirect computing centers into a cloud model for maximum performance and cost effectiveness.

 The simplest model of cloud computing can be created using a pool of servers that have been "virtualized" using an application tool like the Web Services directory, UDDI or a network tool used for server load balancing. This can make multiple servers (real or virtual) appear as a single resource as a cloud. The problem with this model is that it's not a major advancement over existing client/server architectures because it doesn't provide true resource independence i.e. servers are still dedicated to an application set.

The next step in cloud architectures is the use of a software tool that builds cloud-ready applications. Salesforce.com uses this in its Platform-as-a-Service (PaaS) model. The tools ensure that the applications built under the model can be brokered onto multiple servers and will run in a way that doesn't interfere with other users. This model is popular among cloud computing network providers and can be integrated with SaaS offerings from the same providers using their application tools.

Network providers or even enterprises could build a more sophisticated model using a combination of network virtual storage and virtual server technology. This model would allow an application to be created as an "image" or "instance" and stored in the cloud.

When an application request is received, it is assigned to a virtual server, loaded there from stored copy of the application image and granted access to the required data from a storage pool. The use of virtual servers here allows operating system independence, as long as the hardware systems have a common binary execution format. Amazon's Elastic Cloud Computing (EC2) fits this model. A similar model can be created using Java Virtual Machine technology. Java applications can run on any hardware platform, providing even more resource independence.

The "ultimate" cloud computing model would be one where full resource virtualization is combined with resource brokering logic, that takes account of resources cost, network connectivity, performance requirements and user geography. The IBM/Google cloud computing model meets these requirements and a similar framework could be built combining tools from Cisco and other virtualization vendors.

It's important to note that while most cloud computing architectures and offerings are built on a Web Service or SOA framework, the connection to either is not essential. In fact, cloud resources can resemble servers in a client/server arrangement. This application is relatively rare because most companies use cloud computing as a means of supporting special application sets therefore, they supplement their own IT resources. Integrating these special ad-hoc applications with in-house applications requires virtual desktop tools. It performs best and most flexible if all applications are present their capabilities as Web services.

2.2 Developing Holistic Cloud Computing Reference Model

Cloud computing is one of the next significant stage in the Internet’s evolution, providing the means through which everything from computing power to computing infrastructure, applications, business processes to personal collaboration can be delivered to the user as a service wherever and whenever they need.

The "cloud" in cloud computing can be defined as the set of hardware, networks, storage, services and interfaces that combine to deliver aspects of computing as a service.

Consumer Cloud Computing services has been well established ever since mainstream Internet. Known examples are Webmail services and social networking platforms. However the adoption of Cloud Computing within the Enterprise sector has been slow. This slow uptake in Cloud services that promises so much has been primarily influenced by the numerous security risks, concerns and challenges posed within such an environment.

Governance, Risk and Compliance factors of Cloud Services need to be fully assessed and evaluated by organizations to provide informed judgments. Data and Information lifecycle, source and origination, transfer, destination, validation and deletion all need to be understood.

Trans-border data flow across countries with different cyber law jurisdictions need to be carefully considered and any sensitive information leakage resulting in litigation requires the involvement of cyber law legal teams. Periodic rights for 3rd party audit clause, frequent reporting mechanisms of security violations and a clearly defined service level agreement between an organization and the Cloud Service Provider needs to be developed.

With Cloud providers utilizing shared pool of resources, virtualization and isolation, capabilities need to be questioned along with identity access control and management frameworks. Encryption key lifecycle of virtualized environments, portability of information, if the organization decides to move to another Cloud provider are just some critical factors to consider.

2.2.1 Cloud Computing: Taking an Holistic View

The speed at which Cloud computing has planted its feet firmly in the center stage of the IT and Communications world is quite interesting. It's not a technology play in fact it may be hard to identify a single piece of new technology that is fundamental to cloud. And unlike Twitter or Facebook it's not a social-psychology phenomenon in any real sense there is no "man-in-the-street' movement that is driving the uptake or need for cloud computing. It is one of those rare beasts a practical, common-sense driven initiative.

Putting it simply, cloud computing makes much more efficient use of resources. In the early stages these resources are essentially processing power and storage but increasingly the focus of cloud will converge on efficient use of software resources from a bewildering array of sources.  The concept of a user being able to gain access to and pay for these resources on a per-use basis makes great economic sense for everyone from the lone mobile game developer in his garage, to the large scale financial institution.

Unless someone spots a fatal flaw with the concept, over the next ten years we will move from a predominantly distributed computing and storage world to a centralized computing and storage world. Major investment in new levels of security is inevitable and company analysts predict that cloud's road to success will hit the odd pothole around the area of security.

What makes cloud so interesting is that every one of the global vertical industries (Telecoms, Financials, Retail, etc) have two conversations about the cloud, first one how do we become a cloud user to enable more efficient operations and second one, how do we leverage our existing platform assets to become a cloud provider.

2.2.2 Towards holistic cloud management

Despite significant attention and substantial efforts both in industry and academia, cloud computing has not yet reached its full potential. Commonly stated obstacles for cloud adoption include confusion about terminology with multiple delivery models (SaaS, PaaS, IaaS) and deployment scenarios (public clouds, private clouds, etc.). Other frequent concerns relate to risk of outsourcing, data legislation issues, inability to assess trust in external providers, etc. In addition to these obstacles to cloud computing as a concept, there are also technological thresholds in today's cloud offerings, making service provisioning a tedious process.

We summarize the analysis of cloud computing obstacles in five high-level research challenges:

Service life cycle optimization for improved construction, deployment and operation of cloud services.

Adaptive self-preservation with increased autonomy in resource management, enabling fewer administrators to handle increasingly larger systems.

Cloud and service self-management based on non-functional management criteria, extending beyond typical cost-performance tradeoffs to also incorporate aspects of trust, risk and eco-efficiency in decision making.

Support for service deployment and operation in multiple cloud architectures such as private clouds, cloud federation, cross-cloud service provisioning and resource mediation by third-party cloud brokers.

Market and legislative issues including identification of new market roles and business models for clouds, as well as investigation of legal aspects related to the acquisition, transfer and storage of service data.

2.2.3 Cloud Computing Reference Model

The Cloud computing Reference Model (CC-RM) is developed to facilitate the process of Cloud modeling, deployment planning and architecture. As with Service-oriented architecture (SOA), until OASIS promoted the concept of Reference Model, Reference Architecture and Reference implementation, there was no standardized approach to realize a SOA implementation that could be traced to standards, to a sound reference architecture approach, which both contribute to a successful reference implementation.

The Cloud Computing Reference Model (CC-RM) will establish a Cloud modeling and architecture foundation from which an organization can realistically plan, model, architect and deploy Cloud computing in a pragmatic fashion to address real and pressing business and technical challenges. Cloud should not be treated as a solution looking for a problem but as a collection of Cloud patterns that can be configured to meet a wide array of business and technical requirements.

The Cloud reference model developed is comprised of four supporting models or elements, as described below:

Cloud Enablement Model

The core of the Cloud Computing Reference Model is the Cloud Enablement Model. The Cloud Enablement Model describes the tiers of Cloud computing foundation, enablement and business capabilities provided by Cloud platform and service providers to potential consumers of Cloud-enabled technology and business capabilities. The Cloud enablement model is comprised of the range of Cloud technologies and enablement solutions such that all Cloud patterns can be realized by providers and consumers.

Cloud Deployment Model

The cloud deployment model (CDM) describes the range of Cloud deployment scenarios available to the enterprise internal-private Cloud, external public Cloud, hybrid Cloud and community clouds. These deployment scenarios may be mixed and matched to meet a variety of business use cases and requirements.

Cloud Governance and Operations Model

Describes the governance, security and privacy, operations and support, management and monitoring requirements for cloud computing and ensures that the user has considered all the potential operational risks for adopting Cloud for their enterprises.

Cloud Ecosystem Model

The Cloud Ecosystem model considers the requirements of developing and sustaining a Cloud ecosystem comprised of Cloud providers, Cloud consumers, Cloud intermediaries, as well as the Cloud network and "Cloud dial tone" necessary to ensure the cloud is always there for the user. The Cloud ecosystem also includes the various Cloud enablement technologies and cloud providers and consumers of those Cloud enablement technologies to establish the Cloud ecosystem.

The Cloud Computing Reference Model is comprised of four sub-models, with elements or dimensions within each of the sub-models. The components of this Cloud Computing Reference Model are summarized below:

Cloud Enablement Model (CEM)

The Cloud Enablement Model is a core element of the CC-RM. The Cloud enablement Model helps us to identify and relate various Cloud Enablement technologies and capabilities into various Cloud "patterns" that help solve business and technology requirements. The Cloud Enablement Model has four "tiers" of Cloud Enablement capabilities:

Cloud Virtualization Tier

This Cloud enablement tier focuses on the range of technologies and tools that provide hardware and infrastructure virtualization, computing, storage, network and security virtualization, all as a Cloud foundation upon which the other Cloud enablement tiers are layered. Virtualization technology is fairly mature but in the context of Cloud computing, it is still evolving.

Cloud Operating System Tier

The Cloud Operating System Tier (OS) focuses on the class of Cloud technologies that enable virtualization of resources as Cloud enabled capabilities. As such, the Cloud OS tier must provide provisioning, billing and metering, load balancing, resource management, monitoring and management, workflow and "orchestration" of Cloud-enabled resources into pattern-based solutions and more. The Cloud OS tier is what enables higher order Cloud patterns and solutions to be created.

Cloud Platform Tier

The Cloud Platform Tier is comprised of the tools and technologies that enable platforms to be assembled and delivered as a service (PaaS), as well as the individual technology capabilities that enable platforms. These include SOA and Web services, application server and runtime containers, content management, Web and application servers and more. The Cloud platform tier builds on the Virtualization and OS tier to enable application platforms and PaaS capabilities to be built, delivered and consumed as a service.

Cloud Business Tier

The Cloud business tier consists of a broad range of business capabilities and business solutions that are designed and provisioned to consumers as services via the Cloud. While SalesForce CRM and Gmail are common examples, the general use case is broader than the typical software as a service (SaaS) construct offered by analysts. Any end-user solution or capability provided as a service via a Cloud-enabled delivery model fits the Cloud Business Tier. For example, data as a service, knowledge as a service, processes as a service, are all possible Cloud business tier solutions that are within the range of the possible given this construction. The Cloud business tier opens up a broad array of potential Cloud business solutions, far beyond the SaaS convention known to date.

The Cloud Enablement Model is meant to organize the range of Cloud technologies and solutions into sensible tiers that build upon one another, yet enable the complete range of potential Cloud solutions to be imagined, modeled and delivered. Each of the Cloud Enablement Tiers is unique, addresses a targeted class of potential Cloud consumers and support Cloud patterns that can address a broad range of business and technical needs.

Cloud Deployment Model (CDM)

The Cloud deployment model (CDM) is a critical dimension of the CC-RM in conjunction with the Cloud enablement model. The CDM provides an explicit framework for identifying the requirements and differences of various Cloud deployment scenarios. When the Cloud deployment model and the Cloud enablement model are used together, the core of the CC-RM comes to life. These are the core decisions that will have the biggest impact on your Cloud implementation.

The Cloud deployment model and the Cloud Enablement Model are key decisions that determine in many respects the security, architectural and management challenges we will face with Cloud computing. Below are some of the primary Cloud deployment model scenarios.

Private Internal Cloud

Internal clouds or private clouds focus on optimization of internal resources and capabilities leveraging Cloud solutions. Private clouds avoid many of the perceived security risks associated with Cloud, while limiting potential benefits of leveraging external Cloud providers to host our clouds.

Public External Cloud

In public cloud the real benefits of a cloud can be realized in outsourcing, as well as the time taken to market, proven technology solutions. Public cloud service providers, such as Amazon, SalesForce, Google and many other easy-to-access and pre-integrated.

Hybrid Integrated Cloud

This will be the dominant Cloud deployment configuration sooner rather than later. Hybrid Clouds mix the features and capabilities of public and private clouds to address business or mission needs. There is already industry momentum behind defining standard APIs to enable Cloud integration and interoperability, a critical requirement of hybrid Clouds. Hybrid Clouds can offer compelling Cloud patterns that mitigate security risks while taking advantage of public Cloud service offerings.

Community Cloud

A community Cloud deployment is one that is shared by a community of interest (COI) or a group of stakeholders with shared interests. Community Clouds could be established for any number of interest groups, such as industry groups, mission domains in the Department of Defense (DoD), or more specific communities of interest. A community Cloud is public to its members but private to outsiders.

The Cloud Deployment Model and the Cloud Enablement Model are central requirements of the Cloud computing Reference Model. Establishing the parameters of these two sub-models will enable the user to plan and implement Cloud for their enterprise with clarity about the deployment and Cloud enablement patterns necessary to solve your business requirements.

Cloud Governance and Operations Model

The Cloud Governance and Operations Model are in many respects determined after making choices about Cloud Enablement approaches in the context of the Cloud deployment model. Below are the primary elements of the Cloud Governance and Operations Model:

Cloud Governance

Cloud governance is an emerging requirement of Cloud computing and encompasses a broad set of business and technical requirements from the planning and architecture process. It is done through the design-time considerations of Cloud computing, functional and non-functional requirements analysis. On-boarding the enterprise into a cloud (internal, public or hybrid) is successfully deployed when the critical monitoring and operational requirements are satisfied.

Security and Privacy

A hot topic, Cloud security and privacy focuses on challenges of data security, privacy, regulatory compliance for privacy, as well as data integrity, physical and logical security and all other related security requirements for services, applications and interactions in a Cloud ecosystem.

Management and Monitoring

Cloud management and monitoring focuses on the instrumentation and tooling necessary to monitor and manage the Cloud, whether the deployment is an internal private cloud, or whether are leveraging third-party external clouds from Amazon or Salesforce.com, or deploying hybrid integrated Clouds. Either way, we must be able to integrate and automate the monitoring, performance management, alarming and alerting of Cloud events and performance metrics in order to respond to outages, performance degradations and related operational concerns.

Operations and Support

Cloud operations and support focuses on the post-deployment phases of the Cloud Lifecycle Model and is poorly understood in the context of Cloud operations and support. Operations and support for hybrid and public clouds will be a fast-moving area of emphasis and we must spend appropriate time understanding the operations and support requirements based on the Cloud deployment, based on which Cloud enablement tiers and cloud patterns we intend to exploit.

Cloud Ecosystem Model

The last key dimension of the CC-RM is the Cloud Ecosystem Model. The Cloud Ecosystem Model creates and enables the business and operational relationships between Cloud consumers and Cloud providers, as well as the technical and physical connections that enable business operations to be provided and performed over the Cloud. The Cloud Ecosystem is the physical, logical and virtual environment in which Cloud providers, Cloud consumers, cloud solution and technology providers all operate to drive the economic and transactional foundation of Cloud computing as a legitimate business and technology trend. The key elements of the Cloud Ecosystem Model are described below.

Cloud Network/Dial Tone

Cloud dial tone refers to the essential requirement that the network always be available and that Cloud-enabled resources are accessible, or "on."

Cloud Ecosystem Enablement

Cloud ecosystem enablement refers to the capabilities and technologies that allow the Cloud ecosystem to be built and managed, thus connecting Cloud providers and consumers such that they can find one another, operationally engage and conduct business transactions via the Cloud network.

Cloud Consumers and Cloud Providers

The Cloud ecosystem must provide support for all Cloud consumers and providers to engage. There cannot be a Cloud ecosystem unless all stakeholders in the Cloud ecosystem are involved from business and technical consumers to developers, cloud providers, consulting organizations, standards bodies and more.

2.3 Cloud’s System Architecture

Below figures are examples of cloud-based solution architectures where user can build with the RightScale platform using both public and/or private cloud infrastructures.  Most of these architectures can be built using existing Server Templates that are available in the MultiCloud Marketplace.   

Each application is unique and will have a custom set of requirements. The purpose of the system architecture diagrams below are provided real-world examples that can be used as base reference architectures in the cloud.  Once you find a system architecture that is similar to what we are trying to build, we can modify and customize it accordingly to meet our own project's requirements.  The diagrams are designed to demonstrate a particular concept such as disaster recovery or multicloud deployments.  When the user is designing his own solution architectures, he/she should consider integrating several of the concepts described below.

Things to Consider

There are several factors that are needed to take into consideration before designing the cloud-based systems architecture.

Cost: Before we architect site/application and start launching servers, we should clearly understand the SLA and pricing models associated with the cloud infrastructure(s). There are different costs associated with both private and public clouds.  For example, in AWS, data transferred between servers inside of the same datacenter (Availability Zone) is free, whereas communication between servers in different datacenters within the same cloud (EC2 Region) is cheaper than communication between servers in different clouds or on-premise datacenters.

Complexity: Before constructing highly customized hybrid cloud solution architecture, make sure we properly understood the actual requirements of application, SLA, etc.  Simplified architectures will always be easier to design and manage.  A more complex solution should only be used if a simpler version will not suffice.  For example, a system architecture that is distributed across multiple clouds (regions) introduces complexity at the architecture level and may require changes at the application level to be more latency-tolerant and/or be able to communicate with a database that's migrated to a different cloud for failover purposes.

Speed: The cloud gives us more flexibility to control the speed or latency of our site/application.  For example, user could launch different instance types based on his application's needs.  For example, whether he needs an instance type that has high memory or high CPU?  From a geographic point of view which cloud will provide the lowest latency for his users?  Is it necessary or cost effective to use a content distribution network (CDN) or caching service?  For user-intensive applications, the extra latency that results from cross-cloud/region communication may not be acceptable. 

Cloud Portability: Although it might be easier to use one of the cloud provider's tools or services, such as a load balancing or database service, it's important to realize that if and when we need to move from particular tier of architecture to another cloud provider, we will need to modify the architecture accordingly.  

Security: For cloud system architectures, it's important to realize that cross-cloud/region communication is performed over the public Internet and may introduce security concerns that will need to be addressed using some type of data encryption or VPN technology. 

2.3.1 Example Reference Diagrams

The architecture diagrams below shows a progression from simple to more complex reference architectures. 

Single "All-in-one" Server

Use one of the "All-in-one" ServerTemplates, such as the LAMP (Linux, Apache, MySQL, PHP) ServerTemplate to launch a single server that contains a web server (Apache), as well as the application (PHP) and database (MySQL).  

Single Cloud Site Architectures

A description...

Fig 7: Single Cloud Site Architecture

In standard three-tier website architecture, there is at least one dedicated server in each tier of the system architecture.  Figure 7 shows that in the single cloud site architecture, load balancer, application logic, databases and storage are located in the cloud, i.e. Load Balancing Server, Application Server, Database Server. If the user is only testing the interactivity between each tier of his architecture, he may want, to use non-redundant system architecture to save costs and resources. As shown in figure 7, there are dedicated servers for each tier of the application/site which form a non redundant architecture. Usually this kind of architecture is not recommended for production environments.

Redundant 3-Tier Architecture

Any production environment that is launched in the cloud should also have a redundant architecture for failover and recovery purposes.  Typically, we may use a Server Array for our application tier to take advantage of auto-scaling in the cloud, however there may be some scenarios where our application is not designed to auto-scale. In such cases, we can still create a redundant multi-tier architecture where we have redundancy at each tier of reference architecture.  In the figure 8 shown below, there are two load balancer servers, two application servers, as well as master and slave database servers.  A redundant architecture will help protect the site/application from system downtime. 

Figure 8 demonstrates that the use of a striped volume set at the database tier. If the database is large and requires faster backups, we have to consider using a set of striped volumes for data storage.  

A description...

Fig 8: Redundant 3-Tier Architecture

Multi-Datacenter Architecture

If the cloud infrastructure supports multiple datacenters (or zones), it's recommended to spread the system architecture across multiple datacenters to add another layer of redundancy and protection.  Each datacenter in a cloud is designed to be an isolated segment inside the same geographical cloud.  So if a power failure occurs in one datacenter, the other datacenters will be unaffected.  For example, in AWS, each EC2 Region (cloud) may have multiple Availability Zones (datacenters). The benefit of using multiple datacenters is to protect entire site/application from being negatively affected by some type of network/power failure, lack of available resources, or service outage that's specific to a particular datacenter.  

As a best practice it should always leverage multiple datacenters in reference architecture if they are supported by the cloud infrastructure as shown in below figure 9. It shows the multi datacenter architecture, in which it has two data centers, each having a load balancing, application, volume and master database. When on datacenters goes down, then automatically datacenter 2 can be used.

A description...

Fig 9: Multi-Datacenter Architecture

Auto-scaling Architecture

One of the key benefits of the cloud is the ability to horizontally scale (i.e. grow or shrink the number of running server resources) as the demands of your application/site change over time.  With RightScale, user can use Server Arrays to set up a particular tier of architecture to auto-scale based on predefined alert conditions.  Auto-scaling is most commonly used for the application tier for cloud reference architecture. Figure 10 shows the example of auto scaling architecture. It consists of two load-balancers, volume, database and server array where applications are installed. When there is a demand for the application, then it can be added in server array for speedy process and performance improvement.

A description...

Fig 10: Auto Scaling Architecture

2.4 Cloud Deployment Model

2.4.1 Cloud Computing Deployment Models

National Institute of Standards and Technology (NIST) summarized the cloud computing characteristics as on-demand self-service, ubiquitous network access, resource pooling, rapid elasticity and pay per use. The rapid transition towards cloud computing has increased the demands for more deployment models. Selection of these models depends on clients’ data sensitivity and management requirements.

Private Cloud

Private cloud (internal cloud) infrastructure is dedicated to a single particular organization or group. It is not shared with other organizations. Private cloud can be owned or leased. It may be managed by the organization or a third party and can exist at on-premises or off-premises. Private cloud is more expensive and secure when compared to public cloud. Private cloud is hosted inside the organization’s firewall. It can be accessed by users within the organization via intranet as shown in figure 11.

A description...

Fig 11: Private cloud

Private clouds are flexible and service-based. Processes, services and data are managed within the organization. In private cloud there are no additional security regulations, legal requirements or bandwidth limitations that can be present in a public cloud environment, by using a private cloud, the cloud service providers and the clients have optimized control of the infrastructure and improved security, since user’s access and the networks used are restricted.

Public Cloud

Public cloud (external cloud) infrastructure is offered via web applications as well as web services over the internet to the public or a large industry group and is owned by an organization selling cloud services as shown in figure 12. Public cloud provides an elastic, cost-effective way to deploy IT solutions. The term public doesn’t mean that users’ data is publicly visible. Public cloud involves applications such as customer relationship management (CRM), messaging and office productivity. Public cloud providers such as Google or Amazon offer an access control to their clients.

A description...

Fig 12 Public cloud

Hybrid Cloud

This cloud deployment model exists due to mixed needs of an organization. It is combination of two or more cloud service deployment models (Private, Public and Community) as shown in figure 13. Organizations may host critical applications on private cloud and applications with relatively less security concerns on the public cloud.

A description...

Fig. 13 Hybrid cloud

A combination of a public and a private cloud is joined together for the purpose of keeping business-critical data and services in their control on private cloud and outsourcing less-critical processing to the public cloud.

Community Cloud

Community cloud is a shared infrastructure by several organizations and it supports a specific community that has shared concerns e.g., mission, security requirements, policy and compliance considerations. It may be managed by the organizations or a third party and may exist at on-premises or off-premises. Community cloud offers higher level of privacy, security and policy compliances. Examples of community clouds include Google’s "Gov Cloud".

Combined Cloud

Combined cloud is formed of two clouds that have been joined together. It consists of multiple internal and external providers. By integrating multiple cloud services, users will be able to ease the transition to public cloud services while avoiding issues such as Payment Card Industry (PCI) compliance.

Inter Cloud (Cloud of Clouds)

Inter cloud refers to a mesh of clouds that are interconnected based on open standards to provide a universal environment for cloud computing. It is a concept based on the aggregation of deployed clouds similar to internet, which is a network of networks. It refers to an inter-connected global cloud of clouds that provides massive collection of services to the users.

2.4.2 Types of Cloud Deployment Models

There are three types of cloud deployment models available however there are is one another type of cloud deployment model known as community cloud which is being used in some instances. Table 1 list the various cloud deployment models and highlights its characteristics.

Table 1: Cloud Deployment Models

Public Cloud

•Provider Owned &

Managed

•Access by Subscription

•Economic Benefits:

•Reduced IT Service

Delivery cost

• Reduced HW, Systems, software, management and application costs.

Private Cloud

•Client Dedicated

•Access defined by client

•Data Governance rules/regulations

•More Secure

•Economic Benefits:

•Reduced Capex

•Reduced Opex

•Service Level Discipline

Hybrid Cloud

•Enterprise Computing and

private cloud extend outward to consume public compute resource

for peak needs

•Economic Benefits

•Scale private cloud for BAU

•Maintain service levels by scaling externally

•Share cost with vertical with charge back options

Key Patterns:

•User initiated consumption of compute/storage resources through service catalogue self service portal

•Highly scalable and automated provisioning of commodity compute resource

•Pay per use metering & billing

Key Patterns:

•Resource driven

provisioning of

development, test and

production systems

Managing E2E Lifecycle

•Provisioning systems

quickly and application on

boarding with minimal

human effort

Key Patterns:

•Policy driven acquisition of external compute resources beyond the data center firewall

•Shift commodity workloads at periods of peak usage consumption to external compute/storage resources