Unleashing the True Power of Cloud Computing
Integrating systems creates synergy and power that exceeds that of traditional paradigms
By: Bill Kalma
Sep. 30, 2011 07:00 AM
The first generation of cloud computing was revolutionary in that it added business value to organizations by reducing development time, eliminating the need to procure infrastructure, providing massive scaling potential, establishing scale through multi-tenancy and by allowing IT people to focus on solving business problems versus technical ones. As cloud-based applications grew in prominence, they fought to achieve parity with legacy on-premise software solutions that were feature/functionality rich. These goals were achieved in short order, and thus the new generation of cloud solutions was born.
Web 2.0 goes far beyond parity with legacy systems and leverages the power of information sharing, collaboration and the social grid to achieve value that would be impossible with a paradigm of pushing information out from a single location. The new age cloud applications are powerful because they break down silos and simplify the process of converting data into information. As Web 2.0 systems in the consumer space evolve, companies are taking notice and looking to unleash the same power in their enterprise applications.
Enterprises face a unique challenge when adopting a cloud strategy. They do not have the benefit of starting with a blank slate and building their applications from the ground up, but they want to take advantage of the efficiency gains and collaborative benefits. Instead, they must find the best way to harvest the investment they have made in existing databases and applications, while realizing a magnitude of change from the cloud. The secret to Web 2.0 - whether it be the development of new applications from scratch, or the marriage between new and legacy applications - is integration.
Systems integration comes in several forms, each of which has validity based on a particular use case. The goal is to establish awareness of the major types of integration such that they can be appropriately applied as the cloud presents possibilities that we previously thought were impossible. With no one-size-fits-all solution, the synergy between these approaches is where the power is unleashed. For the purposes of this article, the major types of integration are data integration, interoperability and UI mash-ups.
Data Level Integration
First, end-user access to data in the cloud-based system is accomplished through a single secured connection that aids in both performance and complexity. By avoiding the need to make calls to an external system (possible an on-premise one), there is no need for call outs or remote access through a firewall.
An additional benefit is that replicating data maximizes the control that the target system has over the data. The data can be manipulated in the transformation process so that it meets the schema of the target system to ease reporting and data manipulation activity at the application layer. This means that just the right amount of normalizing or de-normalizing of the data can be done at transfer time to meet the application's needs.
There are some inherent disadvantages of replicating data across systems. By definition, the duplication of data requires that the target system and the system of record be synchronized at some interval to close the gap on having "two versions of the truth" exist for end users. In most cases, this is not real time, so each use case needs to be evaluated to balance latency with the overhead of synchronization. By replicating data to a new data store, there is also the need to re-define security rules in the target environment; it would be a design misstep to allow users a workaround by making information available in one environment that is not available in another, but that means replicating security rules. Finally, there is a cost to the development and maintenance of data replication solutions in the form of ETL tools, custom code maintenance, additional storage and, of course, duplicity of data management.
In the cloud space, there are many options for implementing data layer integrations that are not only powerful but also cost-effective when compared to their on-premise counterparts. Some players include Informatica, Cast Iron, Boomi.com, Talend, Pervasive and Scribe.
Integration through interoperability is lightweight and contextual because your cloud application is only accessing the data that it needs from an external service when it needs it. Service calls are made to services to return information to the end user who is shielded from the actual source of the data. For example, suppose a cloud-based application holds a list of customers and it would be useful to aggregate as much data as possible on this customer. One option would be to subscribe to market data, load it into a local data store, and display it as part of your application. With the interoperability of the cloud, it's possible to simply make calls to a third-party NASDAQ, Hoovers, or D&B service to bring back real-time customer data that you don't have to maintain. There are countless services available in the cloud that are more comprehensive and robust than anything we could develop ourselves.
Interoperability is not limited to external services; it can (and should) be leveraged internally as well. In the same cloud-based customer database, it might be useful to aggregate sales information that is sourced in the ERP system. Instead of duplicating the sales history in the cloud, a preferred solution would be to expose sales history data through a service that can be called to return contextual sales information for a customer. From an end-user standpoint, the experience is seamless because the data is displayed to them in the context of the customer application, but there is no replication, synchronization issues or latency.
The potential of interoperability goes far beyond these simple examples. In fact, it is interoperability that is helping to define Web 2.0. Social networking sites like LinkedIn, Facebook and Twitter have services that are consumed in countless other applications to provide information about people, their networks and personal preferences. Furthermore, some of the fast-growing and most popular services today like Groupon, Living Social and Foursquare leverage interoperability with other services to market to individuals rather than demographics, and have unleashed the potential of crowd-sourcing.
The significance of the mashup comes down to reuse and end-user experience. There is a benefit to the cloud developer to be able to tap into and reuse other services (even user interfaces) because doing so drastically reduces development time. It is also a benefit to the end user to work within a familiar context that they already enjoy.
Reader Feedback: Page 1 of 1
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week