KVM: VMs with a data volume already on a reserved device id still cannot start after the upgrade to the device id fix

未关闭
  • 哪里不对？
#14,514 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 1 天内回复

还没有人认领这个 Issue。

评估

难度
4/5
预计耗时
3-5 天
新手友好度
48/100
Issue 类型
缺陷
描述清晰度
基本清楚
活跃度
活跃
技术栈
java
领域
backend, infrastructure

调研方向

Start by tracing the KVM VM start path that assigns device IDs to data volumes, CD-ROMs, and the config drive ISO; the issue does not name source files or tests. Compare how existing volumes with reserved IDs are handled during startup with the detach and reattach behavior described here. Done when affected q35 VMs start after upgrade without manual volume reattachment.

由索引模型根据 Issue 内容生成。

描述

bug component:kvm component:storage component:upgrade
Problem

The device id fixes on 4.22 and main reserve the device ids of the CD-ROMs and the config drive ISO for new data volumes. Volumes that already use one of these device ids (created on an earlier version) are not changed, as noted in both fixes. On q35 these VMs still fail to start with XML error: target '...' duplicated after the upgrade, until the volume is detached and attached again. These VMs could not start before the upgrade either (on 4.23.0.0 a config drive VM with a data volume on 5 also fails with vm.iso.max.count=2), so this is not a regression of the fixes.

Versions

The versions of ACS, hypervisors, storage, network etc..The VMs already fail on the old version (4.22.1.0, 4.23.0.0). They still fail after the upgrade to a build that includes the fix:

  • 4.22.1.0 upgraded to the 4.22 build with the device id fix (4.22.2.0-SNAPSHOT): q35 config drive VM with a data volume on deviceid 4 (target 'sde' duplicated).
  • 4.23.0.0 upgraded to main (24.0.0-SNAPSHOT): q35 VMs with a data volume on deviceid 4, with and without config drive (target 'sde' duplicated). A q35 config drive VM with a data volume on deviceid 5 starts with vm.iso.max.count=1 and fails with vm.iso.max.count=2 (target 'sdf' duplicated).

Environment:

KVM on Oracle Linux 9, qemu-kvm 7.2.0, libvirt 9.0.0, machine type pc-q35-rhel9.2.0 (template settings rootDiskController=scsi, kvm.guest.os.machine.type=pc-q35-rhel9.2.0, UEFI=LEGACY), host.cdrom.max.count 2. NFS primary storage.

The steps to reproduce the bug
  1. On the earlier version (4.22.1.0 or 4.23.0.0), create an L2 network with DefaultL2NetworkOfferingConfigDrive.
  2. Deploy a q35 scsi VM on it (stopped) and attach 3 data volumes. The 3rd gets deviceid 4.
  3. Upgrade the management server and the KVM agents to the version with the fix.
  4. Start the VM. It fails with 530 "Unable to orchestrate the start of VM instance", and the management server log has "Unable to start VM on Host ... XML error: target 'sde' duplicated" for each host.
  5. Detach the volume on deviceid 4 and attach it again: it gets 5 (4.22) or 6 (main), and the VM starts.
What to do about it?

Expected: after the upgrade, VMs with data volumes on the reserved device ids start without a manual detach and attach.

主要语言
Java
星标
3.1k
派生
1.4k
平均合并
7 天 16 小时
30 天内合并 PR
15

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。

apache/cloudstack 的其他 Issue

查看 apache/cloudstack 的全部 Issue

相似的 Issue

更多 Java Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。