Quarkus Batch — Jakarta Batch (JBeret)

Overview

Expert guidance for batch processing in Quarkus 3.36.0 using the Jakarta Batch specification, implemented by JBeret. Provides chunk-oriented (read-process-write) and tasklet-style (batchlet) processing with CDI integration, checkpointing, partitioning, and job management via JobOperator.

Key Capabilities:

• Job definitions via XML (src/main/resources/META-INF/batch-jobs/*.xml) or programmatic API
• Chunk-oriented steps: ItemReader → ItemProcessor → ItemWriter with configurable commit intervals
• Batchlet steps for single-task operations (file moves, SQL procedures, etc.)
• Checkpoint / restart: jobs resume from the last checkpoint on failure
• Job parameters passed at launch time
• Partition: PartitionMapper + PartitionReducer for parallel processing across partitions
• JobOperator API to start, stop, restart, and query jobs
• @Named + @Dependent for CDI-managed batch artifacts
• Integration with @Scheduled for cron-triggered batch jobs
• @QuarkusTest + JobOperator for integration testing

When to Use