redpanda-data/redpanda v25.3.1

on GitHub

Features

• Shadow Linking, for built-in Disaster Recovery, by @michael-redpanda in #27349 and others
• Support for GCP BigLake Metastore as an Iceberg REST Catalog for Iceberg Topics, by @nvartolomei in #27251
• Schema Registry now supports IMPORT mode for schema migration, by @pgellert in #27010
• Connected Client Monitoring
  • Added a cluster connections list command by @graham-rp in #28157
  • Implement cluster-wide API endpoints for monitoring client connections by @pgellert in #28137, and others
• KIP-937 (improved message timestamp validation)
  • Support message.timestamp.{before/after}.max.ms, by @WillemKauf in #27419
  • Adds a new cluster config kafka_produce_batch_validation which controls the level of timestamp validation performed on batches in the redpanda produce path, by @WillemKauf in #27529
• KIP-516 (topic identifiers) by @BenPope in #26968
• KIP-709 (fetch the offsets of multiple consumer groups in a single request) by @BenPope in #27794
• Iceberg: Redpanda now supports protobuf Struct fields in Iceberg tables. A Struct field will be translated into a string column, and its value with be the JSON serialization of the Struct value. by @wdberkeley in #26930
• Iceberg: Adds a configurable DLQ table suffix to support Iceberg catalogs/query engines that don't allow tilde characters and hence the default <topic name>~dlq format for DLQs. by @nvartolomei in #28242
• Iceberg: support schema merging, gracefully retaining information about historical columns. by @nvartolomei in #27202
• Introduces a 'security report' API GET /v1/security/report (Admin API) for monitoring the status of enabled security features by @IoannisRP in #27173
• Ability to repair clusters that have lost controller quorum through a controller 'force reconfigure' API by @joe-redpanda in #27995
• Export of consumer group information via rpk group describe ... --formatted json as json or yaml using the --formatted flag. by @alextreichler in #27647
• Control the global upper bound on message size via kafka_max_message_size_upper_limit cluster property. This property can be used to limit the max value of the max.message.bytes topic property. by @ballard26 in #27997
• Introduce a new "sasl_mechanisms_override" configuration that allows per-listener configuration of available sasl mechanisms. by @IoannisRP in #27458
• Enterprise features "core_balancing_continuous" and "partition_autobalancing_mode" will be turned on by default on a new cluster or when any Enterprise license is applied. by @IoannisRP in #27824