Oracle Integration Cloud: Building Integrations That Survive Production · Module 8 · Error handling, faults and retries
The dead letter, and who reads it
Lesson 85 of 175 · 2 min
Somewhere for records that cannot be processed now. And a queue nobody monitors is a silent data loss with extra steps. That second sentence is why this is a lesson. A dead letter that nobody reads is worse than no dead letter at all: without one, a failing record produces a visible failure; with an unmonitored one, it produces a successful run and a row in a table nobody has opened since it was built. WHAT A USABLE DEAD LETTER CONTAINS: the record itself, enough context to understand it (which run, which file, which step), the failure reason in words, and a timestamp. A queue holding payloads with no reason is a queue nobody can triage, and it fills up. WHO READS IT is the design question, and it has to be a named person with a schedule. Daily for anything financial, weekly for the rest. "The team will check…
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 8 · Error handling, faults and retries
- 1The four kinds of failure
- 2Faults versus errorsFree preview
- 3Scope and fault handlers
- 4What to do with a caught fault
- 5Retry: how many, how far apart
- 6The dead letter, and who reads it
- 7Partial failure in a batch
- 8Compensation, and its limits
- 9Alerting a human
- 10Resubmission: doing it safely
- 11Errors that are not failures
- 12What breaks: five error-handling failures
- 13Lab briefing · Break it five ways