[Bug] FE build cannot resolve org.apache.doris:je:18.3.14-doris-SNAPSHOT: the snapshot is gone from repository.apache.org
维护者通常 1 天内回复
还没有人认领这个 Issue。
评估
- 难度
- 2/5
- 预计耗时
- 1-3 小时
- 新手友好度
- 72/100
- Issue 类型
- 缺陷
- 描述清晰度
- 描述清楚
- 活跃度
- 活跃
- 技术栈
- java
- 领域
- build-system
调研方向
过时的版本由 fe/pom.xml 中的 je.version 属性固定,fe-core 将其作为依赖解析。针对空的本地仓库运行报告中的两条 mvn dependency:get 命令进行复现,然后将失败的坐标 18.3.14 与已发布的 18.3.17 进行比较。当干净的 FE 构建能够解析 je 构件且 catalog 单元测试通过时即为完成,修复方式可以是升级版本,也可以是切换到已发布的 je 版本。
由索引模型根据 Issue 内容生成。
描述
Search before asking
- I had searched in the issues and found no similar issues.
Version
master, branch-4.1 and branch-4.0 (all pin <je.version>18.3.14-doris-SNAPSHOT</je.version> in fe/pom.xml). Observed on a 4.1.5 checkout with an unchanged fe/pom.xml.
What's Wrong?
Since 2026-11-02 a Frontend build with an empty (or expired) local Maven repository fails to resolve org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT:
[ERROR] Failed to execute goal on project fe-core: Could not resolve dependencies for project org.apache.doris:fe-core:jar:1.2-SNAPSHOT
[ERROR] dependency: org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT (compile)
[ERROR] Could not find artifact org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT in central (https://repo.maven.apache.org/maven2)
[ERROR] Could not find artifact org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT in huawei-obs-sdk (https://repo.huaweicloud.com/repository/maven/huaweicloudsdk/)
[ERROR] Could not find artifact org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT in snapshots (https://repository.apache.org/content/repositories/snapshots/)
[ERROR] Could not find artifact org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT in apache.snapshots (https://repository.apache.org/snapshots)
The artifact only ever existed in the Apache snapshot repository. That repository keeps a single snapshot version per artifact: https://repository.apache.org/content/repositories/snapshots/org/apache/doris/je/ now lists only 18.3.17-doris-SNAPSHOT (uploaded 2026-11-02 06:36 UTC), and 18.3.14-doris-SNAPSHOT/maven-metadata.xml returns 404. je is not published anywhere else, including Maven Central.
Machines and CI runners that built before still have the jar in ~/.m2, so the breakage only shows on a fresh environment. That is probably why it has not been noticed yet.
What You Expected?
fe/pom.xml resolves on a clean machine, either by pinning the snapshot that is currently published (18.3.17-doris-SNAPSHOT resolves today) or, better, by depending on a released, non-snapshot build of je so this cannot repeat the next time the snapshot rotates.
How to Reproduce?
With an empty local repository, no project needed:
cd "$(mktemp -d)"
mvn -B -Dmaven.repo.local="$PWD/m2" dependency:get \
-Dartifact=org.apache.doris:je:18.3.14-doris-SNAPSHOT \
-DremoteRepositories=apache.snapshots::::https://repository.apache.org/content/repositories/snapshots/
# -> Could not find artifact org.apache.doris:je:jar:18.3.14-doris-SNAPSHOT in apache.snapshots
mvn -B -Dmaven.repo.local="$PWD/m2" dependency:get \
-Dartifact=org.apache.doris:je:18.3.17-doris-SNAPSHOT \
-DremoteRepositories=apache.snapshots::::https://repository.apache.org/content/repositories/snapshots/
# -> succeeds
Or clone the repository on a machine that has never built Doris and run the FE build.
Anything Else?
Bumping je.version to 18.3.17-doris-SNAPSHOT builds and passes FE unit tests here (checked the org/apache/doris/catalog/**/*Test pattern, 477 tests), so a one-line bump unblocks fresh builds immediately. A released version would be the durable fix.
Are you willing to submit PR?
- Yes I am willing to submit a PR!
Code of Conduct
- I agree to follow this project's Code of Conduct
- 主要语言
- Java
- 星标
- 16.2k
- 派生
- 4k
- 平均合并
- 1 天 17 小时
- 30 天内合并 PR
- 660
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue,再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库,在一个分支上完成修改。
- 提交 Pull Request,并在描述里引用这个 Issue 编号。
apache/doris 的其他 Issue
-
难度 2/5 1-3 小时 新手友好度 82/100
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 65/100
维护者通常 1 天内回复
-
[Bug] array_sort's second comparator argument takes the first argument's name: a nested lambda reads the wrong argument, and alias functions sort nothing可能已有人在做 @adarshomrey 今天认领。 未关闭
难度 2/5 1-3 小时 新手友好度 82/100
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 78/100
维护者通常 1 天内回复
-
[Bug] Upper-case hex in cloud initial rowset id is mis-parsed by BE from_hex (lossy, non-unique RowsetId)可能已有人在做 @siddiqueirshad 于 8 天前认领。 未关闭
难度 2/5 1-3 小时 新手友好度 78/100
维护者通常 1 天内回复
相似的 Issue
-
难度 2/5 1-3 小时 新手友好度 78/100
scylladb/scylla-cdc-source-connector#402 ·
维护者通常 1 天内回复
-
enhancement
难度 2/5 1-3 小时 新手友好度 74/100
维护者通常 1 天内回复
-
bug
难度 2/5 1-3 小时 新手友好度 64/100
维护者通常 1 天内回复
-
enhancement good first issue
难度 2/5 1-3 小时 新手友好度 72/100
ahmedmmordi/store-app#8 ·
-
jhipster heroku fails on heroku git:remote when reusing an existing app outside a git repository未关闭area: triage theme: undefined
难度 2/5 1-3 小时 新手友好度 76/100
jhipster/generator-jhipster#36018 ·
维护者通常 1 天内回复