📝 0 新 Issues
✅ 6 关闭
✨ 0 新 PRs
🎉 5 合并
🔥 核心要点
- 🎉 PR #2900 支持Java API复杂数据类型（NestedRow/ROW），增强客户端类型系统
- 🎉 PR #3032 添加Flink过滤器下推集成与完整文档
- 🐛 PR #3053 修复跨poll边界的过期fetch检测问题
- 🚀 PR #3054 优化getNullCounts()返回int[]替代Long[]，提升性能
- 🎉 PR #2728 批量读取后及时释放VectorSchemaRoot buffer
💬 Issue/PR 动态
Issue 讨论
Issue #2900 [fluss-client] Support Complex Data Types on the Java API (NestedRow/ROW)
- @XuQianJin-Stars: 感谢详细评审，将重构PojoArrayToFlussArray预编译转换器、修复类型不安全问题和调整测试位置
Issue #2992 [test] Unstable test RemoteLogScannerITCase.testScanFromRemoteAndProject
- @github-actions[bot]: 报告RemoteLogScannerITCase测试失败，出现AssertionFailedError
Issue #3048 [lake][iceberg] Iceberg does not support union read for primary key table
- @MehulBatra: 当前架构下主键表union读会触发重型merge-sort join，正通过Deletion vectors优化设计
Issue #3054 [common] Optimize getNullCounts() to return int[] instead of Long[]
- @platinumhamburg: 感谢修复编译问题，代码已合并
PR Review
PR #3032 [flink] Add Flink filter pushdown integration and documentation
- @wuchong: 建议添加验证检查拒绝无效logRecordBatchFilter配置、重构统计列获取逻辑、修复测试数据重复写入问题
PR #3026 [client] Replace Netty PooledByteBufAllocator with ChunkedAllocationManager for Arrow memory allocation
- @Copilot: 指出ChunkedFactory未关闭导致内存泄漏、需要对齐大小跟踪、添加输入验证和单元测试
- @wuchong: 建议移动类到row.arrow.memory包、修复空buffer处理、添加单元测试和JavaDoc改进
PR #2803 [client] Netty prefer heap memory
- @wuchong: 指出配置项未实际使用、建议从Configuration获取preferHeap而非硬编码、需添加单元测试
- @Copilot: 指出API破坏性变更、建议保留旧工厂方法、LengthFieldBasedFrameDecoder子类化方案和注释更新
PR #3053 Fix stale fetch detection across poll boundaries
- @Copilot: 建议日志同时输出fetchOffset和nextFetchOffset以避免误导，修正测试注释中的bucket引用
PR #3057 [spark] Fix column projection on log/upsert read path
- @Yohahaha: 建议将公共函数移动到FlussPartitionReader类中
📝 新建 Issue/PR
Issues
暂无
Pull Requests
暂无
✅ 关闭 Issue/PR
已关闭 Issues
- #3052 [client] LogFetchCollector uses immutable fetchOffset for staleness check, causing cross-poll CompletedFetch to be drained @platinumhamburg
- #3031 [flink] Add Flink filter pushdown integration and documentation @platinumhamburg
- #3021 Optimize DefaultLogRecordBatchStatistics.getNullCounts() to return int[] instead of Long[] @platinumhamburg
- #2726 Clear VectorSchemaRoot to release buffer as soon as possible after a batch read finined. @loserwang1024
- #2258 Support tier map type for iceberg @luoyuxia
- #1707 Iceberg support union read for primary key table in batch mode @beryllw
已合并 PRs
- #3054 [common] Optimize getNullCounts() to return int[] instead of Long[] @platinumhamburg
- #3053 Fix stale fetch detection across poll boundaries @platinumhamburg
- #3032 [flink] Add Flink filter pushdown integration and documentation @platinumhamburg
- #2900 [fluss-client] Support Complex Data Types on the Java API (NestedRow/ROW) @XuQianJin-Stars
- #2728 [client] Clear VectorSchemaRoot to release buffer as soon as possible after a batch read finished. @loserwang1024