v24.3.11

[piyushredpanda]

Bug Fixes

• Fixes a bug in which cached readers for segments undergoing adjacent compaction would not be evicted from the readers_cache. by @WillemKauf in #25825
• Fixes a crash that could happen when an error occurred during translation in an Iceberg Topic. by @andrwng in #25906
• Reduce log level of "Truncate partition to offset" in DeleteRecords API to debug to improve user experience under certain workloads by @michael-redpanda in #25864
• PR #25442 [v24.3.x] archival: Disable cross-term compaction by @Lazin
• PR #25804 [v24.3.x] inventory: Fix ODR violation by @Lazin
• PR #25854 [v24.3.x] apply snapshot in background apply fiber if required by @mmaslankaprv
• PR #25900 [v24.3.x] storage: check for exceptions in log_manager::disk_usage() (Manual backport) by @WillemKauf
• PR #25901 [v24.3.x] datalake: shutdown datalake managers before stopping partitions by @mmaslankaprv

Improvements

• Introduces two new loggers, pandaproxy/requests and schemaregistry/requests for the printing of full request and response bodies of Rest Proxy (Pandaproxy) and Schema Registry requests when enabled at the trace level. by @pgellert in #25879
• Provide a warning level log message when encountering an oversized batch during produce. Also provide error message in produce response. by @michael-redpanda in #25821
• Schema Registry: Improve diagnostics when POST a schema with missing references by @BenPope in #25884
• #25813 The Schema Registry service now uses a new logger named schemaregistry, separating it from the existing pandaproxy logger used for the REST proxy. by @pgellert in #25839
• give state machine implementer a control over the initial recovery policy by @mmaslankaprv in #25910
• PR #24849 [v24.3.x] cloud_storage: Improve logging in the partition_manifest by @Lazin
• PR #25878 [v24.3.x] tests/write_caching: limit number of messages sent in write caching test by @mmaslankaprv

Full Changelog: v24.3.10...v24.3.11

---

This discussion was created from the release v24.3.11.