change log | getting started | documentation

Why?

In (mostly) every kafka consumer we use in our pipeline, the same guarantees are needed: do not move forward until record is successfully consumed (handled without exception), infinitely retry to process failed records and automatically recover. In consequence, we want to manage and commit offsets manually. That's why this abstraction over java consumer API was written.

Features

• support for different consume protocols
• safe consume, retry if exception is thrown
• no need to manage offsets manually
• pluggable tracers for logs and metrics

Usage

Add dependency to project.clj:

See getting started for more info.