Bug Fixes
- Fixes a bug in Redpanda Iceberg manifest deserialization that would cause Redpanda to fail to commit new Iceberg files following an external operation that wrote manifests with the
lower_boundsfield. by @andrwng in #27519 - Fixes a bug where concurrent log level setting requests could lead to a situation where the logger ended up expiring to the incorrect log level. by @pgellert in #27567
- Prevent oversized allocations when the kafka logger is at trace level by @rockwotj in #27816
- Redpanda will now properly clean up discarded on-disk files on certain error cases in the Iceberg topics write path. by @andrwng in #27766
Improvements
- Allows waiting for collection to complete when running a remote-bundle by @JFlath in #27399
- Pandaproxy: Allow more parallel requests per user when HTTP basic authentication is enabled by @pgellert in #27820
- Prevents an oversized allocation in
raftheartbeats. by @WillemKauf in #27777 - Prevents an oversized allocation in the partition balancing subsystem. by @WillemKauf in #27761
- Redpanda will now log less frequently to warn about using the default Iceberg partition spec with ÅWS Glue. by @andrwng in #27744
- Redpanda will now log trace-level messages about failed requests to the Iceberg REST catalog. by @andrwng in #27744
Full Changelog: v25.1.11...v25.1.12