KIP-1349 — Bytes-based configurable snapshot frequency for share groups
Discussion BrokerConsumer
Replaces the records-based share-group snapshot trigger with byte-based thresholds: a broker-level share.coordinator.max.record.bytes.between.snapshots defaulting to 20 MB and a per-group share.max.record.bytes.between.snapshots dynamic config, deprecating share.coordinator.snapshot.update.records.per.snapshot. Counting update records is workload-dependent — the same 500-record default spans roughly 75 KB of __share_group_state on measured workloads but can be orders of magnitude more elsewhere — making disk usage and replay cost unpredictable across share groups sharing one state topic.
Details
| Author | Muralidhar Basani |
| Status | Discussion |
| JIRA | KAFKA-20070 |
| Wiki | View on Apache Wiki |
| Created | 2026-05-17 |
| Last Modified | 2026-09-01 |
Explore how this KIP affects the Kafka protocol in the Protocol Explorer, or see the full KIP database.