Oracle Integration Cloud: Building Integrations That Survive Production · Module 6 · Orchestration patterns
Loops over collections
Lesson 62 of 175 · 2 min
Sequential versus parallel, and the failure semantics of each when record 200 of 500 fails. That specific question is the design decision. SEQUENTIAL processes one at a time, in order. Predictable, easy to reason about, and slow at volume. When record 200 fails, the first 199 are done and the remaining 300 have not started — which is recoverable if you know where it stopped, and a disaster if you do not. PARALLEL processes several at once. Much faster, and when record 200 fails you do not know which others completed without checking, because they were in flight together. Recovery is a reconciliation rather than a restart point. THE QUESTION THAT DECIDES IT IS NOT SPEED, IT IS WHAT HAPPENS ON FAILURE. Ask: does the whole batch have to succeed or fail together? If yes, neither loop gives you that — a loop is not a transaction, and pretending otherwise…
The full lesson is part of the course
The video, the complete written lesson and the module quiz are included in Oracle Integration Cloud: Building Integrations That Survive Production, with a certificate on completion and a fourteen-day refund window.
In this module: Module 6 · Orchestration patterns
- 1App-driven versus scheduled
- 2Invokes, and the shape of a flow
- 3Switch, and branching honestly
- 4Loops over collections
- 5Local variables and assignment
- 6Sub-integrations and reuse
- 7Synchronous versus asynchronous
- 8Stitching in a database or file step
- 9When an orchestration is too big
- 10What breaks: three orchestration failures
- 11Lab briefing · Build a four-step flow