High throughput Kafka clusters + OSS'ing Kafka-Kit

I wrote about gigabyte/sec. throughput, petabyte-scale (on all NVMe) Kafka cluster operations on the Datadog blog. Also, we've open-sourced our tools, Kafka-Kit, which includes a partition mapper with storage bin-packing and a dynamic replication throttle rate control daemon.

Read about it here: Introducing Kafka-Kit: Tools for Scaling Kafka.