19 min read

(For more resources related to this topic, see here.)

What is cloud computing?

If you have been in the IT industry for some time, you probably know what cloud means. For the rest, it is used as a metaphor for the worldwide network or the Internet. Computing normally indicates the use of computer hardware and software. Combining these two terms, we get a simple definition—use of computer resources over the Internet (as a service). In other words, when the computing is delegated to resources available over the Internet, we get what is called cloud computing. As Wikipedia defines it:

Cloud computing is the delivery of computing as a service rather than a product, whereby shared resources, software, and information are provided to computers and other devices as a utility (like the electricity grid) over a network (typically the Internet).

Still confused? A simple example will help clarify it. Say you are managing the IT department of an organization, where you are responsible for purchasing hardware and software (licenses) for your employees and making sure they have the right resources to do their jobs. Whenever there is a new hire, you need to go through all the purchase formalities once again to get your user the necessary resources. Soon this turns out to be a nightmare of managing all your software licenses! Now, what if you could find an alternative where you host an application on the Web, which your users can access through their browsers and interact with it? You are freed from maintaining individual licenses and maintaining high-end hardware at the user machines. Voila, we just discovered cloud computing!

Cloud computing is the logical conclusion drawn from observing the drawbacks of in-house solutions. The trend is now picking up and is quickly replacing the onpremise software application delivery models that are accompanied with high costs of managing data centers, hardware, and software. All users pay for is the quantum of the services that they use. That is why it’s sometimes also known as utility-based computing, as the corresponding payment is resource usage based.

Chances are that even before you ever heard of this term, you had been using it unknowingly. Have you ever used hosted e-mail services such as Yahoo, Hotmail, or Gmail where you accessed all of their services through the browser instead of an e-mail client on your computer? Now that is a typical example of cloud computing.

Anything that is offered as a service (aaS) is usually considered in the realm of cloud computing. Everything in the cloud means no hardware, no software, so no maintenance and that is what the biggest advantage is. Different types of services that are most prominently delivered on the cloud are as follows:

  • Infrastructure as a service (IaaS)

  • Platform as a service (PaaS)

  • Software as a service (SaaS)

Infrastructure as a service (IaaS)

Sometimes referred to hardware as a service, infrastructure as a service offers the IT infrastructure, which includes servers, routers, storages, firewalls, computing resources, and so on, in physical or virtualized forms as a service. Users can subscribe to these services and pay on the basis of need and usage. The key player in this domain is Amazon.com, with EC2 and S3 as examples of typical IaaS. Elastic Cloud Computing (EC2) is a web service that provides resizable computing capacity in the cloud. Computing resources can be scaled up or down within minutes, allowing users to pay for the actual capacity being used. Similarly, S3 is an online storage web service offered by Amazon, which provides 99.999999999 percent durability and 99.99 percent availability of objects over a given year and stores arbitrary objects (computer files) up to 5 terabytes in size!

Platform as a service (PaaS)

PaaS provides the infrastructure for development of software applications. Accessed over the cloud, it sits between IaaS and SaaS where it hides the complexities of dealing with underlying hardware and software. It is an application-centric approach that allows developers to focus more on business applications rather than infrastructure-level issues. Developers no longer have to worry about the server upgrades, scalability, load balancing, service availability, and other infrastructure hassles, as these are delegated to the platform vendors. Paas allows development of custom applications by providing the appropriate building blocks and the necessary infrastructure available as a service.

An excellent example, in this category, is the Force.com platform, which is a game changer in the aaS, specially in the PaaS domain. It exposes a proprietary application development platform, which is woven around a relational database. It stands at a higher level than another key player in this domain, Google App Engine, which supports scalable web application development in Java and Python on the appropriate application server stack, but does not provide equivalent robust proprietary components or the building blocks as Force.com.

Another popular choice (or perhaps not) is Microsoft’s application platform called Widows Azure, which can be used to build websites (developed in ASP.NET, PHP, Node.JS), provision virtual machines, and provide cloud services (containers of hosted applications).

A limitation with applications built on these platforms is the quota limits, or the strategy to prohibit the monopolization of the shared resources in the multitenant environment. Some developers see this as a restriction, which allows them to build applications with limited capability, but we reckon this as an opportunity to build highly efficient solutions to work within governor limits, while still maintaining the business process sanctity.

Specificcally for the Force.com platform, some people consider shortage of skilled resources as a possible limitation, but we think the learning curve is steep on this platform and an experienced resource can pick proprietary languages pretty quickly, average ramp up time spanning anywhere from 15 to 30 days!

Software as a service (SaaS)

The opposite end of IaaS is SaaS. Business applications are offered as services over the Internet to users who don’t have to go through the complex custom application development and implementation cycles. They also don’t invest upfront on the IT infrastructure or maintain their software with regular upgrades. All this is taken care of by the SaaS vendors. These business applications normally provide the customization capability to accommodate specific business needs such as user interfaces, business workflows, and so on. Some good examples in this category are the Salesforce.com CRM system and Google Apps services.

What is Force.com?

Force.com is a natural progression from Salesforce.com, which was started as a sales force automation system offered as a service (SaaS). The need to go beyond the initially offered customizable CRM application and develop custom-based solutions, resulted in a radical shift of cloud delivery model from SaaS to PaaS. The technology that powers Salesforce CRM, whose design fulfills all the prerequisites of being a cloud application, is now available for developing enterprise-level applications.

An independent study of the Force.com platform concluded that compared to the traditional Java-based application development platform, development with the Force.com platform is almost five times faster, with about a 40 percent smaller overall project cost and better quality due to rapid prototyping during the requirement gathering—thanks to the declarative aspect of the Force.com development—and less testing due to proven code re-use.

What empowers Force.com?

Why is Force.com application development so successful? Primarily because of its key architectural features, discussed in the following sections.

Multitenancy

Multitenancy is a concept that is the opposite of single-tenancy. In the Cloud Computing jargon, a customer or an organization is referred to as tenant. The various downsides and cost inefficiencies of single-tenant models are overcame by the multitenant model. A multitenant application caters to multiple organizations, each working in its own isolated virtual environment called org and sharing a single physical instance and version of the application hosted on the Force.com infrastructure. It is isolated because although the infrastructure is shared, every customer’s data, customizations, and code remain secure and insulated from other customers.

Multitenant applications run on a single physical instance and version of the application, providing the same robust infrastructure to all their customers. This also means freedom from upfront costs, ongoing upgrades, and maintenance costs. The test methods written by the customers on respective orgs ensure more than 75 percent code coverage and thus help Salesforce.com in regression testing of the Force.com upgrades, releases, and patches. The same is difficult to even visualize with an in-house software application development.

Metadata

What drives the multitenant applications on Force.com? Nothing else but the metadata-driven architecture of the platform! Think about the following:

  • The platform allows all tenants to coexist at the same time

  • Tenants can extend the standard common object model without affecting others

  • Tenants’ data is kept isolated from others in a shared database

  • The platform customizes the interface and business logic without disrupting the services for others

  • The platform’s codebase can be upgraded to offer new features without affecting the tenants’ customizations

  • The platform scales up with rising demands and new customers

To meet all the listed challenges, Force.com has been built upon a metadata-driven architecture, where the runtime engine generates application components from the metadata. All customizations to the standard platform for each tenant are stored in the form of metadata, thus keeping the core Force.com application and the client customizations distinctly separate, making it possible to upgrade the core without affecting the metadata. The core Force.com application comprises the application data and the metadata describing the base application, thus forming three layers sitting on top of each other in a common database, with the runtime engine interpreting all these and rendering the final output in the client browser.

As metadata is a virtual representation of the application components and customizations of the standard platform, the statically compiled Force.com application’s runtime engine is highly optimized for dynamic metadata access and advanced caching techniques to produce remarkable application response times.

Understanding the Force.com stack

A white paper giving an excellent explanation of the Force.com stack has been published. It describes various layers of technologies and services that make up the platform. We will also cover it here briefly. The application stack is shown in the following diagram:

Infrastructure as a service

Infrastructure is the first layer of the stack on top of which other services function. It acts as the foundation for securely and reliably delivering the cloud applications developed by the customers as well as the core Salesforce CRM applications. It powers more than 200 million transactions per day and more than 1.5 million subscribers. The highly managed data centers provide unparalleled redundancy with near-real-time replication, world class security at physical, network, host, data transmission, and database levels, and excellent design to scale both vertically and horizontally.

Database as a service

The powerful and reliable data persistence layer in the Force.com stack is known as the Force.com database. It sits on top of the infrastructure and provides the majority of the Force.com platform capabilities. The declarative web interface allows user to create objects and fields generating the native application UI around them. Users can also define relationships between objects, create validation rules to ensure data integrity, track history on certain fields, create formula fields to logically derive new data values, create fine-grained security access with the point and click operations, and all of this without writing a single line of code or even worrying about the database backup, tuning, upgrade, and scalability issues!

As compared with the relational database, it is similar in the sense that the object (a data instance) and fields are analogous to tables and columns, and Force.com relationships are similar to the referential integrity constraints in a relation DB. But unlike physically separate tables with dedicated storage, Force.com objects are maintained as a set of metadata interpreted on the fly by the runtime engine and all of the application data is stored in a set of a few large database tables. This data is represented as virtual records based on the interpretation of tenants’ customizations stored as metadata.

Integration as a service

Integration as a service utilizes the underlying Force.com database layer and provides the platform’s integration capabilities through the open-standards-based web services API. In today’s world, most organizations have their applications developed on disparate platforms, which have to work in conjunction to correctly represent and support their internal business processes. Customers’ existing applications can connect with Force.com through the SOAP or REST web services to access data and create mashups to combine data from multiple sources. The Force.com platform also allows native applications to integrate with third-party web services through callouts to include information from external systems in organizations’ business processes.

These integration capabilities of the platform through API (for example, Bulk API, Chatter API, Metadata API, Apex REST API, Apex SOAP API, Streaming API, and so on) can be used by developers to build custom integration solutions to both produce and consume web services. Accordingly, it’s been leveraged by many third parties such as Informatica, Cast Iron, Talend, and so on, to create prepackaged connectors for applications and systems such as Outlook, Lotus Notes, SAP, Oracle Financials, and so on. It also allows clouds such as Facebook, Google, and Amazon to talk to each other and build useful mashups.

The integration ability is the key for developing mobile applications for various device platforms, which solely rely on the web services exposed by the Force.com platform.

Logic as a service

A development platform has to have the capability to create business processes involving complex logic. The Force.com platform oversimplifies this task to automate a company’s business processes and requirements. The platform logic features can be utilized by both developers and business analysts to build smart database applications that help increase user productivity, improve data quality, automate manual processes, and adapt quickly to changing requirements.

The platform allows creating the business logic either through a declarative interface in the form of workflow rules, approval processes, required and unique fields, formula fields, validation rules, or in an advanced form by writing triggers and classes in the platform’s programming language—Apex—to achieve greater levels of flexibility, which help define any kind of functionality and business requirement that otherwise may not be possible through the point and click operations.

User interface as a service

The user interface of platform applications can be created and customized by either of the two approaches. The Force.com builder application, an interface based on point-and-click/drag-and-drop, allows users to build page layouts that are interpreted from the data model and validation rules with user defined customizations, define custom application components, create application navigation structures through tabs, and define customizable reports and user-specific views.

For more complex pages and tighter control over the presentation layer, a platform allows users to build custom user interfaces through a technology called Visualforce (VF), which is based on the XML markup tags. The custom VF pages may or may not adopt the standard look and feel based on the stylesheet applied and present data returned from the controller or the logic layer in the structured format.

The Visualforce interfaces are either public, private, or a mix of the two. Private interfaces require users to log in to the system before they can access resources, whereas public interfaces, called sites, can be made available on the Internet to anonymous users.

Development as a service

This a set of features that allow developers to utilize traditional practices for building cloud applications. These features include the following:

  • Force.com Metadata API: Lets developers push changes directly into the XML files describing the organization’s customizations and acts as an alternative to platform’s interface to manage applications

  • IDE (Integrated Development Environment): A powerful client application built on the Eclipse platform, allowing programmers to code, compile, test, package, and deploy applications

  • A development sandbox: A separate application environment for development, quality assurance, and training of programmers

  • Code Share: A service for users around the globe to collaborate on development, testing, and deployment of the cloud applications

Force.com also allows online browser based development providing code assist functionality, repository search, debugging, and so on, thus eliminating the need of a local machine specific IDE.

DaaS expands the Cloud Computing development process to include external tools such as integrated development environments, source control systems, and batch scripts to facilitate developments and deployments.

Force.com AppExchange

This is a cloud marketplace (accessible at http://appexchange.salesforce.com/) that helps commercial application vendors to publish their custom development applications as packages and then reach out to potential customers who can install them on their orgs with merely a button click through the web interface, without going through the hassles of software installation and configuration. Here, you may find good apps that provide functionality, that are not available in Salesforce, or which may require some heavy duty custom development if carried out on-premises!

Introduction to governor limits

Any introduction to Force.com is incomplete without a mention of governor limits. By nature, all multitenant architecture based applications such as Force.com have to have a mechanism that does not allow the code to abuse the shared resources so that other tenants in the infrastructure remain unaffected. In the Force.com world, it is the Apex runtime engine that takes care of such malicious code by enforcing runtime limits (called governor limits) in almost all areas of programming on the Force.com platform.

If these governor limits had not been in place, even the simplest code, such as an endless loop, would consume enough resources to disrupt the service to the other users of the system, as they all share the same physical infrastructure. The concept of governor limits is not just limited to Force.com, but extends to all SaaS/PaaS applications, such as Google App Engine, and is critical for making the cloud-based development platform stable.

This concept may prove to be very painful for some people, but there is a key logic to it. The platform enforces the best practices so that the application is practically usable and makes an optimal usage of resources, keeping the code well under governor limits. So the longer you work on Force.com, the more you become familiar with these limits, the more stable your code becomes over time, and the easier it becomes to work around these limits.

In one of the forthcoming chapters, we will discover how to work with these governor limits and not against them, and also talk about ways to work around them, if required.

Salesforce environments

An environment is a set of resources, physical or logical, that let users build, test, deploy, and use applications. In the traditional development model, one would expect to have application servers, web servers, databases, and their costly provisioning and configuration. But in the Force.com paradigm, all that’s needed is a computer and an Internet connection to immediately get started to build and test a SaaS application.

An environment, or a virtual or logical instance of the Force.com infrastructure and platform, is also called an organization or just org, which is provisioned in the cloud on demand. It has the following characteristics:

  • Used for development, testing, and/or production

  • Contains data and customizations

  • Based on the edition containing specific functionality, objects, storage, and limits

  • Certain restricted functionalities, such as the multicurrency feature (which is not available by default), can be enabled on demand

  • All environments are accessible through a web browser

There are broadly three types of environments available for developing, testing, and deploying applications:

  • Production environments: The Salesforce.com environments that have active paying users accessing the business critical data.

  • Development environments: These environments are used strictly for the development and testing applications with data that is not business critical, without affecting production environment. Developer environments are of two types:

    • Developer Edition: This is a free, full-featured copy of the Enterprise Edition, with less storage and users. It allows users to create packaged applications suitable for any Salesforce production environment. It can be of two types:

      • Regular Developer Edition: This is a regular DE org whose sign up is free and the user can register for any number of DE orgs. This is suitable when you want to develop managed packages for distribution through AppExchange or Trialforce, when you are working with an edition where sandbox is not available, or if you just want to explore the Force.com platform for free.

      • Partner Developer Edition: This is a regular DE org but with more storage, features, and licenses. This is suitable when you expect a larger team to work who need a bigger environment to test the application against a larger real-life dataset. Note that this org can only be created with the Salesforce Consulting partners or Force.com ISV.

    • Sandbox: This is nearly an identical copy of the production environment available to Enterprise or Unlimited Edition customers, and can contain data and/or customizations. This is suitable when developing applications for production environments only with no plans to distribute applications commercially through AppExchange or Trialforce, or if you want to test the beta-managed packages. Note that sandboxes are completely isolated from your Salesforce production organization, so operations you perform in your sandboxes do not affect your Salesforce production organization, and vice versa. Types of sandboxes are as follows:

      • Full copy sandbox: Nearly an identical copy of the production environment, including data and customizations

      • Configuration-only sandbox: Contains only configurations and not data from the production environment

      • Developer sandbox: Same as Configuration-only sandbox but with less storage

  • Test environments: These can be either production or developer environments, used speficially for testing application functionality before deploying to production or releasing to customers. These environments are suitable when you want to test applications in production such as environments with more users and storage to run real-life tests.

Summary

This article talked about the basic concepts of cloud computing. The key takeaway items from this article are the explanations of the different types of cloud-based services such as IaaS, SaaS, and PaaS. We introduced the Force.com platform and its key architectural features that power the platform types, such as multitenant and metadata. We briefly covered the application stack—technology and services layers—that makes up the Force.com platform. We gave an overview of governor limits without going too much detail about their use. We discussed situations where adopting cloud computing may be beneficial. We also discussed the guidelines that help you decide whether your software project should be developed on the Force.com platform or not. Last, but not least, we discussed various environments available to developers and business users and their characteristics and usage.

Resources for Article :


Further resources on this subject:


LEAVE A REPLY

Please enter your comment!
Please enter your name here