Lagforge
Model Kafka partitions, consumer concurrency, lag burn-down, retention, replication, and failure recovery without a live cluster.
Focused field tools
Apply partitions checks to the recorded evidence, document assumptions, and retain a concrete verification result before changing a live system.
Apply keys checks to the recorded evidence, document assumptions, and retain a concrete verification result before changing a live system.
Apply consumers checks to the recorded evidence, document assumptions, and retain a concrete verification result before changing a live system.
Apply lag checks to the recorded evidence, document assumptions, and retain a concrete verification result before changing a live system.
Apply throughput checks to the recorded evidence, document assumptions, and retain a concrete verification result before changing a live system.
Apply retention checks to the recorded evidence, document assumptions, and retain a concrete verification result before changing a live system.
Lagforge is a deterministic planning model, not a Kafka client. Broker limits, key distribution, rebalances, transactions, compression, storage, and application behavior can change outcomes. Validate against the target cluster.