A Day in a Life of a Java Developer in Amadeus
summary
Amadeus replaces a single threaded in‑house batch with Spring Batch + Quartz to handle millions of daily records, adding reusable readers/writers, chunked parallelism, skip/retry and DB-stored schedules. Multi-tenant setup uses one database with many schemas, passing the target schema in job parameters, setting a thread-local schema in a custom JobLauncher and a DataSource that overrides getConnection to call setSchema. Workarounds cover JobRepository startup checks, separate thread pools and backward compatibility; future options include Azure Batch and standalone cloud jobs.
AI-written summary · about
chapters (12)
- 0:00Start
- 1:26Amadeus in a few words
- 1:47Four decades of technological evolutions
- 4:52What Was The New Idea?
- 9:19Quartz Spring Batch
- 10:02Why Quartz?
- 12:18What Are The Challenges?
- 14:38What is Multi-Tenant Architecture?
- 15:56Spring Batch In Single-Tenant
- 17:38Spring Batch In Multi-Tenant Architecture
- 20:44Spring Batch In Action
- 34:08Future Vision
related talks
01h 24mJava Batch JSR-352
44mBatching for the Modern Enterprise
49mBatch Processing in Action
01h 00mScaling Batch Applications
01h 07mJBatch or Not-So-Big Data
01h 01mSpring Batch 5.2 and Beyond: What’s New and What’s Coming Next
46mAsynchronous Job Processing in Java with JobRunr
06mManaging a Java Batch job with JobStreamer