📝 5 新 Issues
✅ 4 关闭
✨ 7 新 PRs
🎉 6 合并
💬 Issue/PR 动态
Issue 讨论
Issue #3586 [ci] Elixir client integration unit tests failed
- @swuferhong: 请求 fresh-borzoni 和 loserwang1024 查看 Rust CI 失败问题，后因新失败用例重新打开此 issue
- @beryllw: 定位问题根因为 rustler 版本不匹配：Rust 端已升级到 0.38.0，但 Elixir 端仍锁定 0.37.3
Issue #3540 [fs] Support AWS S3 credentials provider mode
- @litiliu: 已处理评审意见，添加 S3 凭证提供者文档并重新基于 main 分支
Issue #3562 Fast fail due to the duplicate write results for tiering committer
- @luoyuxia: 询问为何 enumerator 会为同一表重新分配新一轮，提及遇到类似超时问题
Issue #3569 [tiering] introduce the option of `lake.tiering.io.tmpdir` to avoid disk exhaustion
- @zuston: @luoyuxia 请查看此 PR
Issue #3512 [docs] Add Apache Gravitino/Iceberg Rest catalog quick-start guide for Fluss-Iceberg integration
- @MehulBatra: 询问任务进展，表示如有需要可提供帮助
- @VladBanar: 表示还有 2 天休假结束，届时将创建 PR，为延迟致歉
Issue #3588 [rust][tests] Add DynamicWriteBatchSizeEstimator integration tests
- @Prajwal-banakar: 将 PR 从 fluss-rust 仓库迁移过来，已处理评审意见
Issue #3523 [lake/paimon] Support custom Paimon path
- @wzx140: 根据建议将 paimon.path 校验移至 PaimonLakeCatalog.alterTable 中实现
Issue #3492 [FIP-37] Add bitmap scalar functions and register via FlussCatalog
- @platinumhamburg: 确认评审通过，邀请 polytos 和 wuchong 再看一眼
Issue #3549 [lake/hudi][docs] Add Hudi tiering service documentation
- @luoyuxia: 提示需要 rebase 到 main 分支，因为文档已重组
Issue #3233 [server] Support altering table.log.ttl table option
- @luoyuxia: @swuferhong 请查看此 PR
Issue #3465 [test] Unstable test FlinkUnionReadPrimaryKeyTableITCase.testUnionReadInStreamMode
- @github-actions[bot]: 报告测试失败：bucket 未同步的 AssertionError
Issue #3598 [ci] Cache only elixir deps, not _build, to fix stale NIF load
- @fresh-borzoni: @swuferhong 这个修复应该对你有帮助，请查看
PR Review
PR #3140 [FIP-43][client] Support log scanner and writer for multiple tables
- @fxbing: 提出多个代码评审意见：ScanRecords 是否应返回 progress-only 结果、readAsTargetSchema 是否仍需要、是否需要添加每表指标、LogFetcher 清理逻辑等
PR #3540 [fs] Support AWS S3 credentials provider mode
- @litiliu: 回复评审意见：添加 DynamicTemporaryAWSCredentialsProvider 拒绝检查、保持懒加载解析、使用 Fluss 内部 StringUtils 替代 commons-lang3、添加 IRSA 文档说明
PR #3222 [server] Respect the option of `table.log.ttl` for local segments
- @gyang94: 建议在 LogTieringTask.run() 开始时冻结时间戳，以便各步骤使用相同时间值
- @zuston: 同意建议，表示明天查看
PR #3539 [server] Reject KV leader promotion under disk write lock
- @swuferhong: 更新重试逻辑复用共享协调器调度器，为后续迁移添加 TODO 标记
📝 新建 Issue/PR
Issues
- #3599 [ci] Elixir integration flaky: fluss_nif.so fails to load @fresh-borzoni
- #3595 Support ALTER MATERIALIZED TABLE ... AS in FlinkCatalog @xuyangzhong
- #3593 Raise disk write-limit ratio in FlussClusterExtension to avoid test flakiness @xuyangzhong
- #3592 Add fluss-flink-2.1-0.9.1.jar @chentjxxx
- #3586 [ci] Elixir client integration unit tests failed @swuferhong
Pull Requests
- #3601 [lake]Tolerate Iceberg lake table existent if the schema and properties matches @Guosmilesmile
- #3600 [rust] chore: bump Arrow stack to 59 @michaelmitchell-bit
- #3596 [flink] Support ALTER MATERIALIZED TABLE ... AS in FlinkCatalog @xuyangzhong
- #3591 [flink-action][common][server] Support KV shared SST orphan file cleanup @platinumhamburg
- #3590 [flink] Clean orphan children under unknown databases @platinumhamburg
- #3589 [flink] Support remove_orphan_files action for Flink 1.18 @platinumhamburg
- #3588 [rust][tests] Add DynamicWriteBatchSizeEstimator integration tests @Prajwal-banakar
✅ 关闭 Issue/PR
已关闭 Issues
- #3599 [ci] Elixir integration flaky: fluss_nif.so fails to load @fresh-borzoni
- #3593 Raise disk write-limit ratio in FlussClusterExtension to avoid test flakiness @xuyangzhong
- #3586 [ci] Elixir client integration unit tests failed @swuferhong
- #2358 [DOC] Restructure documentation to highlight Streaming Lakehouse deployment and core features @luoyuxia
已合并 PRs
- #3598 [ci] Cache only elixir deps, not _build, to fix stale NIF load @fresh-borzoni
- #3597 [server] Limit max bucket num per table partition @swuferhong
- #3594 [test] Raise disk write-limit ratio in FlussClusterExtension to avoid test flakiness @xuyangzhong
- #3587 [rust/elixir] Align rustler version to 0.38.0 in mix.exs and mix.lock @beryllw
- #3567 [common] Limit default max bucket count to 20000 @swuferhong
- #2699 [docs] Reorganize Streaming Lakehouse documentation to higlight deployment and core features @MehulBatra