Overview (Design Patterns)

In the last 20 years, organizations have recognized that data is a valuable core business asset and the more data stored, refined, aggregated and analyzed, then the more valuable are the insights it can provide. However, the main issue is that it can often be difficult to access, orchestrate, and interpret this data.

As application systems have become more complex and distributed, the movement and communication of that data between applications has become more important. When data is moving across systems, it isn’t always in a standard format and the role of data integration makes data agnostic.

Principle of Application Orchestration

To have a better understanding of the data integration, we need to be aware of the principle of Application or Service Orchestration. This represents the scenario of integrating two or more applications and/or services together to automate a process, or synchronize data in real-time.

Often, point-to-point integration may be used as the path of least resistance. However, point-to-point integration always leads to a complex tangle of application dependencies (often referred to as "spaghetti code") that is very hard to manage, monitor and maintain.

The utilization of principles of Orchestration provides the following advantages:

• An approach to integration that decouples applications from each other

• Provide capabilities for message routing, security, transformation and reliability

• Defines a method to manage and monitor your integrations centrally.

The Critical Nature of Data Integration

As the IT world has evolved over time, the manner and methods in which companies run and manage their critical enterprise data has changed. In earlier days of IT, all data integration efforts were performed through individualized tools that were designed and optimized to handle specific types of data (JMS, SOAP, etc).

However, now organizations now employ numerous tools from a variety of vendors in order to access and retrieve their business intelligence data. For many different reasons (application complexity, system integration, microservices, etc), we now find that our core data is scattered across numerous applications and databases. In our current environment, data integration plays an ever increasing critical role in our standard business processes. Thus, it is pivotal to ensure companies have a solid solution to manage their core data integration needs.

More and more resources are needed to allow data integration to solve the problem of moving, transforming, and consolidating information from various parts of the enterprise (systems, databases, applications, files, and web services). Within these processes, our data must also undergo cleansing, standardizing, de-duplication, manipulation, and synchronization between sources.

Data Integration Problems

Not having a common data integration tool across the enterprise will make it difficult for businesses to maintain and manage information and impede their communication patterns. For example, one department has a different way of managing customer data than another, trying to consolidate data from both departments can be difficult.

Additionally, consolidating our core enterprise data into a unified view (like products or customers) becomes another impediment for businesses that must be overcome. The implementation of extract, transform, and load (ETL) methods to perform Master Data Management (MDM) are common, but are often complex and pose difficulties with flexibility and scalability.

Attempting to integrate many different data integration tools can be an expensive, inefficient and risky proposition for development teams. Additionally, vendor's have a tendency to package integration tools to include numerous other components, resulting in higher costs and bulky solutions.

This can lead to risky investments in time and resources to manage and maintain an infrastructure consisting of a great deal of hardware and software. Additionally, the use of certain data types can call for specific tools, thereby forcing application systems to implement integration from multiple disparate vendors (with all the known headaches associated with this implementation). When a multitude of integration solutions are deployed across the enterprise, it will result in inefficiency due to a lack of consistency and simplicity.

Data Integration Design Patterns

To make our critical enterprise data more usable and to increase its availability faster, developers can use data integration patterns to standardize the integration process. There are five data integration patterns based on business use cases.