SGE Tests segfault in CI
- 主要语言
- Python
- 星标
- 256
- 派生
- 150
- PR 合并指标
- 30 天内没有已合并 PR
描述
Opening an issue to triage the segfault that seems to be happening in th SGE tests.
For some time the SGE tests have been failing. When you look at the logs of [a recent run](https://github.com/dask/dask-jobqueue/actions/runs/10250992141/job/28357989642) on `main` it contains the following error.
```
*** Error in `/opt/anaconda/envs/dask-jobqueue/bin/python3.8': double free or corruption (!prev): 0x00005626c18fa470 ***
/bin/bash: line 1: 588 Aborted (core dumped) pytest /dask-jobqueue/dask_jobqueue --verbose --full-trace -s -E sge
```
I also opened #652 to bump the minimum Python version here to 3.9 and I see a [similar issue happening](https://github.com/dask/dask-jobqueue/actions/runs/10251115360/job/28358406864?pr=652) but with a slightly different error.
```
*** Error in `/opt/anaconda/envs/dask-jobqueue/bin/python3.9': corrupted size vs. prev_size: 0x00005560404681a0 ***
/bin/bash: line 1: 592 Aborted (core dumped) pytest /dask-jobqueue/dask_jobqueue --verbose --full-trace -s -E sge
```
Strangely in both cases `pytest` reports everything has passed.
```
9 passed, 270 skipped in 26.69s
```
贡献指南
调研方向
从链接的 GitHub Actions 运行开始,使用 `pytest /dask-jobqueue/dask_jobqueue --verbose --full-trace -s -E sge` 重现 SGE 套件。检查与 SGE 相关的测试及其 CI 环境,以确定为什么 pytest 表示测试已通过后,Python 却报告分配器损坏。确定原因并使 SGE CI 运行在没有 segfault 的情况下完成,即视为完成。
由索引模型根据 Issue 内容生成。
评估
- 技术栈
- python
- 领域
- distributed-systems, testing
- Issue 类型
- 缺陷
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 活跃度
- 停滞
- 描述清晰度
- 需要澄清
- 新手友好度
- 35/100