Kafka metadata: from hand-rolled to DataHub

Nobody owns a Kafka topic until something breaks. The first answer was a topic-ownership service designed and implemented end to end, aggregating ACLs, log directories and consumer groups into one view of who depends on what.

Hand-rolled metadata has a ceiling, and reaching it led to DataHub. The contributions went upstream into LinkedIn's open source project: ElasticSearch reliability work on the search and indexing path, and integration with the Confluent Schema Registry so schemas and lineage came from the source of truth rather than a copy.