apache / apache/cloudstack

Volumes not destroyed after VM expunging

未关闭
#9,221 3 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看
archive status:needs-reproducing
主要语言
Java
星标
3.1k
派生
1.4k
平均合并
6 天 19 小时
30 天内合并 PR
32

描述

##### ISSUE TYPE

* Bug Report

##### COMPONENT NAME

~~~
core
~~~

##### CLOUDSTACK VERSION

~~~
4.19.0.1
~~~

##### CONFIGURATION

##### OS / ENVIRONMENT

KVM hosts (both Ubuntu 22 and OL8 as OS) with NFS v4 primary storage
Using library: libvirt 8.0.0
Using API: QEMU 8.0.0
Running hypervisor: QEMU 6.2.0

##### SUMMARY

Some volumes (only ROOT volumes) remains in Destroy state after the related VM is correctly expunged. Trying to manually delete them results in the following error: "cannot unlock file 'path-to-the-qcow2': Permission denied. Libvirtd run as root on the hosts and I found no permission/ownership issues on the qcow2 disks.

##### STEPS TO REPRODUCE

~~~
I'm not really sure on how to reproduce the issue, seems related only to root disks in our environnement, as data disks give no issues if I try to delete them.
~~~

##### EXPECTED RESULTS

~~~
When the VM is expunged also the root disk is deleted as well
~~~

##### ACTUAL RESULTS

~~~
The root disk remain in the UI in "Destroy" state
~~~

贡献指南

打开贡献指南

调研方向

该 issue 未指明源文件、测试或入口点。首先跟踪报告的 KVM/NFS 环境中清除 VM 时的核心卷清理流程,并调查根卷为何仍处于 Destroy 状态;完成的标准是根卷随已清除的 VM 一起被移除,且不再出现报告的权限错误。

由索引模型根据 Issue 内容生成。

评估

技术栈
java
领域
cloud, infrastructure
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
25/100

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。