Add Kafka source connectivity dry-run
Implemented SeaTunnel's existing source connectivity dry-run SPI for Kafka using runtime-equivalent configuration and bounded AdminClient metadata checks.
apache/seatunnel · #12167
Deployment preflight feature
SeaTunnel can now validate Kafka broker authentication and topic metadata before allocating resources or submitting a synchronization job.
Problem
`--dry-run connect` reported Kafka sources as unsupported, so invalid credentials and missing explicit topics were not detected before job submission. Static configuration and supplied schemas could not establish broker access.
Approach
Validates literal topics, patterns, both multi-table option names, security settings, and runtime output schemas within a shared metadata-request budget. It does not read records, access offsets, create groups, or alter normal source execution.
Impact and scope
- Finds authentication and explicit-topic failures before a Kafka ingestion job consumes cluster resources.
- Supports existing single-table, pattern, and multi-table source configurations through one preflight path.
- Performs a side-effect-free metadata check without creating consumers, producers, groups, topics, or offsets.
Validation
- All 98 Kafka connector tests passed on Java 8 and Java 11, including factory discovery, schema parity, timeouts, interruption, and cleanup.
- Three authenticated Kafka 7.0.9 integration tests passed on each JDK, covering successful checks, invalid credentials, missing topics, and unchanged broker state.
- A SeaTunnel maintainer approved the feature with explicit appreciation; final checks passed and the GitHub-verified merge commit credits Goutam Adwant and closed GH-12166.