Pages on Kafka broker features that change storage cost or consumption patterns, based on the Kafka 4.3 documentation.
Cutting broker disk cost by moving older log segments to object storage with tiered storage: how the local and remote tiers work, choosing a RemoteStorageManager plugin, unsupported compacted topics and the steps to disable it later.
Using Kafka share groups for queue-style workloads where consumers process records individually without one consumer per partition, when they replace a separate message queue, and their status by Kafka version.
How eligible leader replicas make leader election safer when the ISR shrinks: the strict min ISR rule, enabling it on clusters upgraded to 4.0, defaults on new 4.1 clusters, downgrade and checking ELR state.
The availability versus data loss trade-off when every in-sync replica is gone: unclean.leader.election.enable defaults, enabling it dynamically in KRaft, forcing an election with kafka-leader-election and when it is ever justified.
Tell us what you're working on, whether it's a deployment, an audit, a security test or a cyber range. You'll speak with an engineer who can help you scope it.