Features
Bug Fixes
- Fixes a bug that prevented Iceberg uploads from proceeding after uncommitted data was removed by Kafka retention during a period where Iceberg was disabled. by @andrwng in #27417
- Redpanda Iceberg topics will now serialize some Iceberg snapshot summary metrics that are required for certain query engines (e.g. Redshift) to run against Redpanda's tables. by @andrwng in #27413
- Redpanda will now automatically detect when using the Glue Iceberg REST Catalog and avoid using the
purgeRequested=true
option when dropping tables. This would previously result in commits to Iceberg being blocked when an Iceberg topic was deleted and recreated. by @andrwng in #27380 - rpk: Allow using
rpk config status
in dedicated clusters. by @andresaristizabal in #27352
Improvements
- A new configuration option iceberg_topic_name_dot_replacement specifies a string that will replace dots in Iceberg topic names when those names are converted into Iceberg table names. This allows working around downstream systems which do not allow table names with dots in them. Be careful to avoid table name collisions by picking a proper replacement string, and do not change the replacement string if there exists an Iceberg topic with dots in the name. by @wdberkeley in #27320
- PR #27107 [v25.2.x] Data migrations worker rework by @bashtanov
- PR #27361 [v25.2.x] sr/service: do not log shutdown exception with error severity by @mmaslankaprv
- PR #27368 [v25.2.x]
storage
: move location ofmark_segment_as_finished_self_compaction()
by @WillemKauf
Full Changelog: v25.2.2...v25.2.3