[HDFS-17535] I have confirmed the EC corrupt file, can this corrupt file be restored?

• Type: Bug
• Status: Open
• Priority: Blocker
• Resolution: Unresolved
• Affects Version/s: 3.1.0
• Component/s: ec, hdfs
• Assignee: Unassigned
• Reporter: ruiliang
• Created: 24/May/24 08:37
• Updated: 17/Jul/24 02:57

Description

我了解到 EC 确实存在文件损坏的错误
https://issues.apache.org/jira/browse/HDFS-15759

1：我已确认 EC 损坏文件，此损坏文件可以恢复吗？
有重要数据导致我们生产数据丢失问题？有办法恢复吗？
检查 EC 块组：blk_-9223372036361352768
状态：错误，消息：EC 计算结果不匹配。：ip 为 10.12.66.116 块为：-9223372036361352765

2：https://github.com/apache/orc/issues/1939我想知道如果你选择了你当前的代码（GitHub pull request #2869），我可以跳过与HDFS-14768，HDFS-15186， 和HDFS-15240？

hdfs 版本 3.1.0

谢谢

 

Latest findings: It is a machine network problem, the cpu si(soft interrupt) is too high, nn loses dn heartbeat, nn sends to dn to recover and reconstruct.

Because the Weaver-Scope service of k8s is installed on the server, conntrack interruption times out seriously, affecting all network usage.