ipfs/kubo

Ipfs gets into bad state after computer sleep

Open

#2,777 创建于 2016年6月1日

在 GitHub 查看
 (10 评论) (3 反应) (0 负责人)Go (13,906 star) (2,725 fork)batch import
help wantedkind/bugstatus/deferredtopic/libp2p

描述

I suspect this happens after I sleep my laptop and/or change networks. I'll notice that IPFS doesn't work (I can't get content, and content I ipfs add isn't available on other nodes). In the logs, I'll see a message like

19:41:54.117 ERROR ipns-repub: Republisher failed to republish: failed to find any peer in table repub.go:72

The process is still running. I'm running this on a laptop, and using systemd (in user mode) to run ipfs.

I collected debugging information, and uploaded it here: ipfs-debug-20150531-republisher-failed.tar.gz.

贡献者指南

技术栈
go
领域
backend
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。
4
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。
over 1 week
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。
unclear
前置要求
Go programmingIPFS conceptsunderstanding of peer to peer networkingdebugging distributed systems
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。
15
研究方向
The issue describes a problem where IPFS enters a bad state after computer sleep or network changes, with the error 'Republisher failed to republish: failed to find any peer in table' in logs. The debugging tarball (ipfs debug 20150531 republisher failed.tar.gz) may contain helpful logs and state. Investigate the IPNS republisher logic in 'repub.go' and how it handles network transitions. Look for related issues or PRs that discuss systemd integration or sleep/wake handling. Consider testing on a laptop with network changes to reproduce. The scope of a fix is unclear; it may require changes to the republisher's peer discovery or reconnection logic.