check-cassandra-catalog-lock: a refused migrate changed a keyspace that held only the lock row

未关闭 适合新手
#871 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 4 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
84/100
Issue 类型
缺陷
描述清晰度
描述清楚
活跃度
活跃
技术栈
cassandra, rust
领域
databases, testing

调研方向

从 devtools/check-cassandra-catalog-lock 开始,追踪它如何在 migrate --yes 和预览前后读取 keyspace 的表名和 settings 键。更新错误输出,在两次读取结果不同时打印两者。完成标准是将来出现不匹配时能显示前后值;重新运行此检查以确认输出。

由索引模型根据 Issue 内容生成。

描述

backend:cassandra ci:stability

devtools/check-cassandra-catalog-lock failed a keyspace whose settings table holds only the lock row is left as it was in job 127151625065 of cassandra-catalog-checks, on fix/stream-arn-account-check on 5 November 09:07 UTC. That is 1 failure in 168 jobs between 3 and 5 November.

For this case the script creates a keyspace whose settings table has the catalog's columns and no rows. It reads the keyspace's table names and settings keys, then runs migrate --yes and the migrate preview, and reads them again. migrate --yes takes the lock, refuses the keyspace and releases the lock. The preview takes no lock. The two reads differed once. The script prints FAIL without the two reads, so the job log does not show what changed. A lock row that the release did not delete would explain it, but the log cannot confirm that.

A first step is to print both reads when they differ, so the next failure shows its cause.

主要语言
Rust
星标
517
派生
60
平均合并
5 天 3 小时
30 天内合并 PR
79

环境准备

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

ExtendDB/extenddb 的其他 Issue

查看 ExtendDB/extenddb 的全部 Issue

相似的 Issue

更多 Rust Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。