Hybrid architectures that mix on-premise infrastructure with cloud-native services are increasingly the norm, and Kafka is no exception. This article covers connecting Confluent Cloud Kafka to a self-hosted Kafka cluster using Cluster Linking, so topics stay in sync across environments without bolting on a separate replication tool.
What Cluster Linking does
Cluster Linking is a Confluent Cloud feature that replicates topics and partitions directly between clusters — cloud-to-cloud, cloud-to-on-prem, or across regions — without running MirrorMaker 2. It's built for hybrid, multi-region setups.
Why teams replicate across environments
- Disaster recovery — if one cluster goes down, the other keeps serving.
- Global data distribution — replicate topics closer to where they're consumed.
- Hybrid cloud architecture — keep on-prem and cloud Kafka in sync as a long-term architecture.
- Data migration — move from on-prem to Confluent Cloud with zero downtime.
The full Medium post walks through the cluster setup on both sides, enabling the link, and verifying replication end to end.