Building an Integration Process | BOOMI

Once a company logs into their account on the Boomi AtomSphere web site, integration processes can be designed and built using a visual designer that includes access to a library of pre-built connectors and process maps. Using familiar point-and-click, drag-and-drop techniques, users can build very simple to very sophisticated integration processes with exceptional speed. No coding is required.


Since its inception, Boomi has focused on simplifying the creation of integration processes for application integration, data integration, and B2B integration. By identifying the common steps needed to automate complex integration scenarios, a series of common integration components have been created and are available to all Boomi users. When developing an integration process, these components are connected to create an end-to-end integration workflow.







Standard Boomi Integration Components include:
Connector
connect to any application or data source
Always the first and last steps of an integration workflow, the Connector enables access to another application or data source. The connector sends/receives data and converts it into a normalized XML format. A Connector’s primary role is to "operationalize an API" by abstracting the technical details of an API and providing a wizard-based approach to configuring access to the associated application. Connectors are also configurable to capture only new or changed data, based on the last successful run date of an integration process.
Data Transformation
transform data from one format to another
While core to any integration, the data stored in various applications is rarely, if ever, semantically consistent. For example, a Customer represented in one application will have different fields and formats from that of another application. Using Boomi's Data Transformation components, users can map data from one format to another.

Any structured data format is supported, included XML, EDI, flat file, and database formats. While transforming data, the user can also invoke a variety of field-level transformations as well to transform, augment, or compute data fields. Over 50 standard functions are provided. Users can also create their own functions and re-use them in subsequent projects.
Decision
execute business logic and data integrity checks
Boomi’s Decision components enable true/false data validation that enables users to explicitly handle a result based on the programmed logic. For example, an order can be checked against the target system to see if it has already been processed. Based upon the outcome of the data check, the request will be routed down either the 'true' or 'false' path. Other examples include checking products referenced in an invoice to ensure they exist before processing the invoice, etc.
Cleanse
data cleansing and validation
Integrations are only as successful as the quality of data that gets exchanged. Boomi’s Cleanse components allow users to validate and "clean" data on a field-by-field, row-by-row basis to ensure that fields are the right data type, the right length, and the right numeric format (e.g. currency). Users have an option of specifying whether they wish to attempt to auto-repair bad data or simply reject rows that are “dirty”. All validation results are routed through a “clean” or “rejected” path which allows users to explicitly handle either scenario.
Message
user-defined dynamic notifications
For any step of the integration workflow, Message components can be used to create dynamic notifications that make use of content from the actual data being integrated. This allows the creation of messages like "Invoice 1234 was successfully processed for customer ABC Inc." Connectors are then used to deliver the message to the appropriate end point.
Route
dynamic, content based routing
Route components examine any of the content in the actual data being processed or use numerous other properties available to the user (such as directory, file name, etc.) and route the extracted data down specific paths of execution.
Split
intelligent data splitting and aggregation
Split components re-organize data into logical representations, such as business transactions. For example, you can take incoming invoice headers and details, and create logical invoice documents to ensure the applications being integrated process or reject the entire invoice vs. discrete pieces.
Testing Integrations
Real-time
To further simplify and shorten the integration development cycle, users can test their integrations directly in the Build environment. Integration testing includes watching as the data moves through the integration process, viewing the actual data, monitoring exactly where the integration fails, and click-thru on the failed step to examine error messages.





Configuring Integration Automation
Integration processes can be configured for “hands off” execution. AtomSphere provides both real-time as well as batch-style automation options depending on the requirements of the integration, as defined below:

Event-based Invocation - direct Atom invocation.

Included in every Boomi Atom is a lightweight HTTP Server. Data can be HTTP-posted to a specific Atom, and that data will be processed in real-time.
Event-based Invocation - remote Atom invocation.

Boomi also provides a Trigger API, which allows you to securely invoke an integration process that is running inside an Atom, regardless of where that Atom may be running, without opening any holes in the firewall. This is a very powerful option when you wish to provide external access to integration processes or trigger an integration based on some event in another application.
Schedule-based Invocation

Included in every Boomi Atom is a schedule manager, capable of invoking integration processes based on a schedule configured by the user. Invocations can be scheduled to run as frequently as every minute. Schedule-based invocation option requires no changes to external applications.
.