Integration pillar

Enterprise systems rarely exist independent of other components, as most have key dependencies on other systems within an organization such as identity management, data storage, or analytics. Well-architected systems are integrated systems - they build on existing capabilities rather than re-inventing a feature, recognize the strengths of other systems for certain workflows, and align with existing enterprise standards and systems to reduce complexity and encourage secure, supported approaches.

ArcGIS has been designed and architected to support the shared goals of easy integration, broad standards-based interoperability, and a robust set of extensibility options across web, desktop, mobile, and server-based patterns.

Integration often brings together the capabilities and resources of multiple business information systems to amplify organizational productivity. In earlier eras of information technology, integration requirements were commonly deferred as they added scope and cost to development projects. Today, a broadening majority of workers are familiar with integrated consumer technologies.  Paired with financial pressures that organizations face to increasingly handle complex, orchestrated activities in less time and with limited human and technical resources, enterprise integration enables achievement of strategic outcomes by changing the way people work, enabling the creation and delivery of greater business value.  Effective enterprise integration is a hallmark of effective digital transformation.

Integration is a general term that needs specification to be able to proceed to design details. One way to break integration types into common categories is to describe integration as taking place at three different architectural tiers:

• Presentation tier or application integration – This tier is focused on user interaction. Integration at the presentation tier will differ based on the chosen user application type (web, mobile, or desktop). This process typically requires light weight workloads that are responsive for users within a few seconds.
• Service-level integration – The application tier, also, in services-oriented architecture parlance, known as the services tier, is where data are processed and processes are orchestrated. System-to-system interactions occur in this tier and medium-weight workloads, contextual content, and business logic are coordinated between applications and data. Requests are expected to be responsive in the time span of seconds to minutes.
• Data-level integration – The data tier, also referred to as the persistence tier, is where data are managed and stored. Authoritative data stores, databases, file shares, and data pipelines (processes for ingestion, transformation, enrichment, and propagation of digital content) are present in this tier. These processes can be performed across multiple systems, commonly with scripted processes for extraction, transformation, and loading (ETL) of digital content between systems.  Data tier processes typically take minutes to hours to complete.

The Integration Approaches and Methods article introduces additional approaches and options for integrating, which can be useful when approaching a new requirement or system integration as a way to help understand the level of complexity and potential effort involved.