alibaba/GraphScope

Coordinator should not try to pull all resources after restarts, it should connect.

Open

#2,872 创建于 2023年6月13日

在 GitHub 查看
 (0 评论) (0 反应) (0 负责人)HTML (2,401 star) (301 fork)batch import
component:coordinatorgood first issue

描述

In kubernetes deployment if the coordinator restarts due to OOM or pod evicted, it will try to pull all components again upon start up, and it will fail cause all resources is already exists, thus causing a infinite failed condition.

It should try to connect in such case.

贡献者指南

技术栈
gokubernetes
领域
devopsinfrastructure
议题类型
bug
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
3
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
1-3 hours
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
stale
清晰度议题是否清楚说明期望改动、验收标准和下一步。
mostly clear
前置要求
Go programmingKubernetes basics
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
40
研究方向
Examine the coordinator startup logic in the GraphScope Kubernetes deployment, likely in the go files under the coordinator package. Identify where it attempts to pull all resources on restart and modify it to connect to existing resources instead. Look for related comments or code sections handling pod restarts.