Integration & Automation
What you must understand
Boomi integration work revolves around processes, shapes, and documents. Know the lifecycle: a process starts with a start shape (connector, listener, or scheduled trigger), flows through data shapes (map, decision, branch, route), and ends with a stop shape. Understand the difference between process execution (a single run) and process schedule (recurring trigger), and how process properties and document properties differ in scope. Dynamic process properties are runtime-only and vanish after execution; document properties travel with a specific document. Be fluent in try/catch scopes, process call versus subprocess reuse, and how branch shapes evaluate conditions top-down with a default path. Also know connectors: database, disk, FTP, HTTP, and application connectors differ in operation support and batching behavior.
Traps that appear
Questions often test shape ordering and whether a map runs before or after a decision. Another common trap is confusing retry logic on a connector operation with a try/catch scope. Watch for questions about document chunking and how batch sizes affect memory. Also expect traps around process property precedence when the same name exists at multiple levels, and around listener versus polling connectors.
Worked example
You need to process a file, split it into records, and route failures to an error queue. The rule of thumb: split early, map once, and wrap only the risky connector call in try/catch. A decision shape after the split evaluates a field; each branch maps to a target shape. If a record fails, the catch path writes to the error queue and continues, so a single bad record does not abort the batch. Remember that a stop shape inside a branch terminates that path only.
Rule of thumb
Design for idempotency: every process should tolerate a replay without duplicating data.