📝 3 新 Issues
✅ 1 关闭
✨ 4 新 PRs
🎉 2 合并
🔥 核心要点
- 🎉 PR #3528 和 PR #3522 合并，增加了 Hudi Flink union read 的测试基础和集成测试
- 🚀 PR #3529 优化客户端写入性能，避免非分区表每条记录重建 AutoPartitionStrategy
- ✨ PR #3521 提交了对 Flink 2.3 的支持，目前正在解决 CI 问题
- 📚 PR #3515 新增 Apache Polaris/Iceberg REST catalog 快速入门指南，已通过端到端测试验证
- 🔒 PR #3526 增强安全性，拒绝客户端传递的 S3 凭证选项
💬 Issue/PR 动态
Issue 讨论
Issue #3283 Introduce Hudi Compaction
- @luoyuxia: 该 Issue 已通过 PR #3518 关闭
Issue #3476 [lake/paimon] Support custom Paimon lake table path
- @wzx140: 移除了无关修改，缩减了代码量，请求再次审查
Issue #3420 [lake/tiering] Support reporting watermark to Paimon snapshot
- @Shawn-Hx: 解释 Watermark 是实现基于 PartitionMarkDone 的前置条件
- @luoyuxia: 同意 Watermark 必要性，但建议限制 PR 范围避免过度实现
Issue #3521 [flink] add support for flink 2.3
- @sd4324530: 指出 CI 失败与当前 PR 修改无关
Issue #3486 [security] Redact sensitive config values in startup logs
- @litiliu: 请求 fresh-borzoni 协助审查代码
Issue #3515 [docs] Add Apache Polaris/Iceberg REST catalog quick-start guide
- @MehulBatra: 要求确认文档在本地 Docker 环境的端到端可用性
- @GabrielBBaldez: 已在 Docker 上验证成功，修复了 S3FileIO 配置问题
Issue #3396 [lake] Fix DvTableReadableSnapshotRetriever scanning wrong snapshot
- @fresh-borzoni: 询问 PR #3519 合并后是否还需要此 PR
Issue #3529 [client] Avoid per-record AutoPartitionStrategy construction in WriterClient.doSend in log tables
- @binary-signal: 请求审查，并解释了缓存策略在分区表上的性能影响
PR Review
PR #3424 [docs] add docs to integrate Fluss + Iceberg via Flink with AWS Glue and Hive
- @MehulBatra: 建议图表风格保持专业统一，Mermaid 或 draw.io 均可
PR #3519 [lake] Fix DV readable snapshot scanning and premature snapshot deletion
- @fresh-borzoni: 指出旧数据保留逻辑可能导致存储增长，建议优化扫描遍历
PR #3486 [security] Redact sensitive config values in startup logs
- @fresh-borzoni: 指出 GCS 密钥未覆盖、动态配置可能泄露敏感信息，建议统一敏感列表
PR #3529 [client] Avoid per-record AutoPartitionStrategy construction in WriterClient.doSend in log tables
- @morazow: 建议使用 isPartitioned，增加测试和 Javadoc
- @fresh-borzoni: 建议仅在创建分区时获取策略，避免引入缓存字段
- @binary-signal: 已添加测试和注释，已处理评论
📝 新建 Issue/PR
Issues
- #3527 [client] WriterClient.doSend rebuilds AutoPartitionStrategy per record on non partitioned tables @binary-signal
- #3524 [client][filesystem] Reject client-side S3 credential options @litiliu
- #3520 [flink] support flink 2.3 @sd4324530
Pull Requests
- #3529 [client] Avoid per-record AutoPartitionStrategy construction in WriterClient.doSend in log tables @binary-signal
- #3526 [client] Reject client-side S3 credential options @litiliu
- #3523 [lake/paimon] Support custom Paimon path @wzx140
- #3521 [flink] add support for flink 2.3 @sd4324530
✅ 关闭 Issue/PR
已关闭 Issues
- #3283 Introduce Hudi Compaction @fhan688
已合并 PRs
- #3528 [lake/hudi] Add Hudi Flink union read log table IT @fhan688
- #3522 [lake/hudi] Add Flink union read test base for Hudi @fhan688