Kafka stability is hard to achieve, especially in high throughput environments. If you wish to hear about the the challenges of handling Kafka clusters in production you can listen to my interview on the Data Engineering Podcast (https://www.dataengineeringpodcast.com/kafka-troubleshooting-in-production-episode-406?fbclid=IwAR27bYXVK7MS2KNAF3Y2I4Z2wiydhrwi46zLarfChq_GNF-lkHB2dIHpNn8) where I talked about real production issues that can occur in Kafka clusters and how to handle them.
These production issues are also covered in my new book (https://www.amazon.com/Kafka-Troubleshooting-Production-Stabilizing-premises/dp/1484294890) where they’re assembled into a comprehensive troubleshooting guide for Kafka clusters deployed either in the cloud or on-premises.
If you're an SRE, DevOps, DataOps or SysAdmin in charge of maintaining a Kafka cluster up and running, or just interested in better understanding of latency issues in Linux, this book is relevant to you.