📝 11 新 Issues
✅ 5 关闭
✨ 5 新 PRs
🎉 5 合并
🔥 核心要点
- 🎉 PR #2594 修复 ADD COLUMN 后部分更新时的 ArrayIndexOutOfBoundsException
- 🎉 PR #2948 实现客户端零拷贝延迟解析 ByteBuf，避免深拷贝提升性能
- 🎉 PR #2978 修复 PeriodicSnapshotManager.start() 静默禁用快照的问题
- 🐛 Issue #2986 Spark 批量读取 PK 表使用随机列作为主键时失败
- ✨ PR #2995 新增支持 log scanner 扫描为 Arrow record batch
💬 Issue/PR 动态
Issue 讨论
Issue #2007 Add missing authentication for RPC calls
- @LiebingYu: 将 issue 分配给 vaibhavk1992，建议分多个 PR 实现各种认证方法
- @vaibhavk1992: 提出分两阶段实现的方案：外部 RPC 和内部 RPC 的授权检查
Issue #2956 [spark] Add union read for lake-enabled log tables
- @Yohahaha: 感谢 PR，创建了 #2983 跟踪，建议在 PR 描述中添加 closes 关键字
- @fresh-borzoni: 说明此 PR 已实现 log 表 union read，PK 表 union read 后续跟进
- @polyzos: 确认 #2983 已在此 PR 中实现，可关闭重复 issue
Issue #2983 [spark] Batch union read for log table
- @fresh-borzoni: 说明此功能已在 #2956 中实现
- @polyzos: 确认功能已实现，建议关闭重复 issue
Issue #2948 [rpc] Client zero-copy lazy parse ByteBuf to avoid deep memory copy
- @fresh-borzoni: 感谢审核，已处理评论内容
- @polyzos: 请求 rebase 后合并，最终 LGTM 批准
Issue #2594 [BugFix] Fix ArrayIndexOutOfBoundsException in partial updates after ADD COLUMN
- @fresh-borzoni: 感谢审核，已处理评论意见
- @polyzos: LGTM 批准合并，感谢贡献
PR Review
PR #2951 [server] Add server-side filter execution and client-side integration
- @wuchong: 提出多项改进建议：异常处理、代码简化、参数封装、缓存优化等
PR #2976 Introduce RPC Layer for KV Scan
- @fresh-borzoni: 询问序列化问题和异常类的区别，建议添加 @PublicEvolving 注解
- @leekeiabstraction: 建议命名改进，提出拆分 closeScanRequest 为独立请求
- @polyzos: 回复 CI 问题，确认命名约定
PR #2989 [filesystem] Support AssumeRole STS for RustFS
- @Copilot: 指出 STS client 资源泄漏风险、配置验证缺失、测试覆盖不足等问题
- @fresh-borzoni: 请求审核者帮助审查此 PR
PR #2995 [client] Support log scanner scan to arrow record batch
- @Copilot: 指出投影逻辑、内存安全（数组暴露、retain 不平衡）、Allocator 性能和依赖作用域问题
PR #2956 [spark] Add union read for lake-enabled log tables
- @Yohahaha: 建议改进分区合并逻辑、添加注释、优化测试用例
- @fresh-borzoni: 接受建议，处理了评论意见
PR #2972 [flink] Implement SupportsWatermarkPushDown for FlinkTableSource
- @loserwang1024: 询问 uid 生成的用途
- @LiebingYu: 解释 uid 对 Flink 状态管理的重要性，参考 Kafka connector 实现
PR #2718 [flink] Support ALTER TABLE to modify watermark for Fluss tables
- @loserwang1024: 指出 watermark 不是非物理列，建议将计算列修改移至另一 PR
PR #2769 [server] Integrate remote directory selector into table/partition creation
- @LiebingYu: 建议保留部分代码，其他问题将在另一个 PR 中解决
PR #2855 [kv] Support downloading CDC log segments from remote storage during KV table recovery
📝 新建 Issue/PR
Issues
- #2994 [test] Unstable test RemoteLogITCase.remoteLogMiscTest @app/github-actions
- #2993 [test] Unstable test Flink120ProcedureITCase.testDropKvSnapshotLeaseProcedure @app/github-actions
- #2992 [test] Unstable test RemoteLogScannerITCase.testScanFromRemoteAndProject @app/github-actions
- #2990 [docs] Add documentation for lookup.insert-if-not-exists feature @platinumhamburg
- #2988 ArrayIndexOutOfBoundsException in partial updates after ADD COLUMN @fresh-borzoni
- #2986 [spark][bug] Batch read PK table failed when use random column as primary key @Yohahaha
- #2985 [spark] Add spark tiering service @Yohahaha
- #2984 [spark] Batch union read for pk table @Yohahaha
- #2983 [spark] Batch union read for log table @Yohahaha
- #2982 [test] Unstable test CommitRemoteLogManifestITCase.testDeleteOutOfSyncReplicaLogAfterCommit @app/github-actions
- #2979 Paimon sorted reader fails when primary key contains TIMESTAMP @luoyuxia
Pull Requests
- #2995 [client] Support log scanner scan to arrow record batch @luoyuxia
- #2991 [docs] Add Insert If Not Exists section to Lookup Joins documentation @platinumhamburg
- #2989 [filesystem] Support AssumeRole STS for RustFS @fresh-borzoni
- #2987 [spark][bugfix] Use pk index for comparator align with SortMergeReader @Yohahaha
- #2981 [WIP] Support variant type and shredding @swuferhong
✅ 关闭 Issue/PR
已关闭 Issues
- #2988 ArrayIndexOutOfBoundsException in partial updates after ADD COLUMN @fresh-borzoni
- #2979 Paimon sorted reader fails when primary key contains TIMESTAMP @luoyuxia
- #2977 [Server] PeriodicSnapshotManager.start() silently skips scheduling when initialDelay is 0 @platinumhamburg
- #2961 Allow enabling lakehouse for log tables without bucket key created before cluster-level lakehouse is enabled @luoyuxia
- #1184 Client lazy parse ByteBuf to avoid deeply memory copy @swuferhong
已合并 PRs
- #2980 [lake/paimon] Fix Paimon PK comparator row type @luoyuxia
- #2978 [Server] Fix PeriodicSnapshotManager.start() silently disables snapshot @platinumhamburg
- #2973 [server] Allow enabling lakehouse for log tables without bucket key created before cluster-level lakehouse is enabled @luoyuxia
- #2948 [rpc] Client zero-copy lazy parse ByteBuf to avoid deep memory copy @fresh-borzoni
- #2594 [BugFix] Fix ArrayIndexOutOfBoundsException in partial updates after ADD COLUMN @fresh-borzoni