kvm: fix restore-attach picking a CD-ROM slot as the target device (NAS backup) by slavkap · Pull Request #14266 · apache/cloudstack

slavkap · 2026-09-29T13:22:35Z

Description

Problem

Restoring a single volume from backup and hot-attaching it to a running KVM VM (restoreVolumeFromBackupAndAttachToVM) fails, regardless of the volume's storage type (NFS, StorPool, RBD, LINSTOR).

####Root cause

getDeviceToAttachDisk() picks the target device name via:
virsh domblklist --domain <vm> | tail -n 3 | head -n 1 | awk '{print $1}'

This returns the third-from-last line, not the last one — it "works" only by coincidence. Every CloudStack KVM VM has two pre-allocated, empty CD-ROM drives (hdc, hdd, added by #13101) that sort alphabetically ahead of the real (virtio) disks, so the pipeline picks up a CD-ROM (or an already-attached disk, depending on disk count) instead of computing a truly free name. virsh attach-disk then rejects the bogus/duplicate target.

Fix

Add --details to domblklist and filter to Type == "disk" rows before picking the last one, so CD-ROMs are never selected.

Types of changes

Breaking change (fix or feature that would cause existing functionality to change)
New feature (non-breaking change which adds functionality)
Bug fix (non-breaking change which fixes an issue)
Enhancement (improves an existing feature and functionality)
Cleanup (Code refactoring and cleanup, that may add test cases)
Build/CI
Test (unit or integration test code)

Bug Severity

Screenshots (if appropriate):

How Has This Been Tested?

Manually tested with NFS and StorPool as a primary storage

getDeviceToAttachDisk found the next free disk device by piping `virsh domblklist` through `tail -n 3 | head -n 1`, just grabbing a row by position without checking what device it actually was. domblklist sorts targets alphabetically, so a VM's empty IDE cdrom slots can sort after the real disks and get picked up instead — e.g. selecting hdc and incrementing it to hdd, an already-existing IDE target, instead of the next free virtio disk. This made hot-attaching a restored volume to a running VM fail with virsh attach-disk erroring out on the bogus target. This surfaced after apache#13101 ("Multiple CD-ROM / ISO Support Per VM"), which pre-allocates a second empty cdrom slot (hdc, hdd) at boot for every VM. With two cdrom rows now sorting ahead of the VM's disks, the tail/head window shifted enough to land on a cdrom row instead of a real disk — before that change, a single cdrom slot didn't push the window that far. Add --details to domblklist and filter to rows with Type=="disk" before taking the last one, so cdrom slots are excluded and the device name is always derived from an actual disk.

codecov · 2026-09-29T13:31:41Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 19.91%. Comparing base (510d0ec) to head (c737927).

Additional details and impacted files

@@            Coverage Diff            @@
##               main   #14266   +/-   ##
=========================================
  Coverage     19.91%   19.91%           
- Complexity    20194    20200    +6     
=========================================
  Files          6373     6373           
  Lines        577230   577229    -1     
  Branches      70696    70696           
=========================================
+ Hits         114942   114960   +18     
+ Misses       449722   449700   -22     
- Partials      12566    12569    +3

Flag	Coverage Δ
uitests	`3.71% <ø> (ø)`
unittests	`21.18% <100.00%> (+<0.01%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

Copilot

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Copilot review overview

Review effort: Lite
Findings: 1 · 2

Open (3)

The comment claims CD-ROM slots “sort after the real disks”, but the PR description/root cause… · New
Using an array-of-generic (new List[1]) as a mutable holder is not type-safe and typically… · New
This test name suggests it verifies CD-ROM entries are filtered out, but the assertions only check… · New

What changed in this PR

Fixes KVM restore+hot-attach by ensuring the next target device name is computed from actual disks (excluding CD-ROM slots) when parsing virsh domblklist output.

Changes:

Update getDeviceToAttachDisk() to call virsh domblklist --details and filter rows where Device == disk.
Adjust existing unit test to match the updated piped command ordering and awk program.
Add a unit test asserting --details is requested.

File	Description
plugins/hypervisors/kvm/src/main/java/com/cloud/hypervisor/kvm/resource/wrapper/LibvirtRestoreBackupCommandWrapper.java	Fixes device selection logic by filtering `domblklist` output to disk devices only.
plugins/hypervisors/kvm/src/test/java/com/cloud/hypervisor/kvm/resource/wrapper/LibvirtRestoreBackupCommandWrapperTest.java	Updates/extends tests to validate the new `domblklist --details` + awk filtering behavior.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Copilot

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Copilot review overview

Review effort: Lite
Findings: 2 · 1

Open (3)

virsh domblklist --details exposes a Type column, but the awk filter is currently checking… · New
virsh domblklist --details exposes a Type column, but the awk filter is currently checking… · New
This test locks in the same incorrect awk field index as the production code. Once the awk program… · New

Resolved since last review (3)

The comment claims CD-ROM slots “sort after the real disks”, but the PR description/root cause…
This test name suggests it verifies CD-ROM entries are filtered out, but the assertions only check…
Using an array-of-generic (new List[1]) as a mutable holder is not type-safe and typically…

+        // --details adds a Type column so cdrom slots (hdc, hdd) can be filtered out — they sort
+        // alphabetically ahead of virtio disks, so without this the selected row is a cdrom, not
+        // an actual disk.
+        String[] domblkCmd = new String[] { Script.getExecutableAbsolutePath("virsh"), "domblklist", "--domain", vmName, "--details" };


        // "invalid char" and produce no output.
-        String[] awkCmd = new String[] { Script.getExecutableAbsolutePath("awk"), "{print $1}" };
-        Pair<Integer, String> result = Script.executePipedCommands(Arrays.asList(domblkCmd, tailCmd, headCmd, awkCmd), 0);
+        String[] awkCmd = new String[] { Script.getExecutableAbsolutePath("awk"), "$2==\"disk\"{print $3}" };


            // The commands are executed without a shell, so the program must carry no shell quotes.
            Assert.assertEquals("awk", awkCmd[0]);
-            Assert.assertEquals("{print $1}", awkCmd[1]);
+            Assert.assertEquals("$2==\"disk\"{print $3}", awkCmd[1]);


abh1sar · 2026-09-30T16:35:26Z

-        String[] domblkCmd = new String[] { Script.getExecutableAbsolutePath("virsh"), "domblklist", "--domain", vmName };
-        String[] tailCmd = new String[] { Script.getExecutableAbsolutePath("tail"), "-n", "3" };
-        String[] headCmd = new String[] { Script.getExecutableAbsolutePath("head"), "-n", "1" };
+        // --details adds a Type column so cdrom slots (hdc, hdd) can be filtered out — they sort


The comment here is not right. --details doesn't change the ordering in any way:

[root@ref-trl-12455-k-Mr9-abhisar-sinha-kvm1 ~]# virsh domblklist --domain i-2-5-VM --details Type Device Target Source ---------------------------------------------------------------------------------------------------------- file disk vda /mnt/eaae5267-a60a-369d-a4da-a13db16e3aa2/250bd073-3a2c-4875-9ba0-29bf632801b7 file disk vdb /mnt/eaae5267-a60a-369d-a4da-a13db16e3aa2/6f259e43-74a5-461e-b857-e945bcff7b45 file disk vdc /mnt/eaae5267-a60a-369d-a4da-a13db16e3aa2/bca15b84-3acf-4b1d-8466-3866e563c21f file cdrom hdc - file cdrom hdd - [root@ref-trl-12455-k-Mr9-abhisar-sinha-kvm1 ~]# virsh domblklist --domain i-2-5-VM Target Source ------------------------------------------------------------------------------------------ vda /mnt/eaae5267-a60a-369d-a4da-a13db16e3aa2/250bd073-3a2c-4875-9ba0-29bf632801b7 vdb /mnt/eaae5267-a60a-369d-a4da-a13db16e3aa2/6f259e43-74a5-461e-b857-e945bcff7b45 vdc /mnt/eaae5267-a60a-369d-a4da-a13db16e3aa2/bca15b84-3acf-4b1d-8466-3866e563c21f hdc - hdd - [root@ref-trl-12455-k-Mr9-abhisar-sinha-kvm1 ~]#

--details does add the type Device column (not Type) which awk $2==\"disk\" uses to filter by type.

Copilot AI balanced review requested due to automatic review settings September 29, 2026 13:22

boring-cyborg Bot added the component:kvm label Sep 29, 2026

Copilot AI reviewed Sep 29, 2026

View reviewed changes

Copilot started reviewing on behalf of slavkap September 29, 2026 14:02 View session

address copilot comments

c737927

Copilot AI review requested due to automatic review settings September 29, 2026 15:26

Copilot AI reviewed Sep 29, 2026

View reviewed changes

Copilot started reviewing on behalf of slavkap September 29, 2026 15:53 View session

abh1sar reviewed Sep 30, 2026

View reviewed changes

Navigation Menu

kvm: fix restore-attach picking a CD-ROM slot as the target device (NAS backup) - #14266

kvm: fix restore-attach picking a CD-ROM slot as the target device (NAS backup)#14266
slavkap wants to merge 2 commits into
apache:mainfrom
storpool:fix-restore-and-attach-datadisk

slavkap commented Sep 29, 2026

Uh oh!

codecov Bot commented Sep 29, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

abh1sar Sep 30, 2026

Uh oh!

This branch has not been deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

slavkap commented Sep 29, 2026

Description

Problem

Fix

Types of changes

Bug Severity

Screenshots (if appropriate):

How Has This Been Tested?

Uh oh!

codecov Bot commented Sep 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Copilot review overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Copilot review overview

Uh oh!

abh1sar Sep 30, 2026

Choose a reason for hiding this comment

Uh oh!

This branch has not been deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

codecov Bot commented Sep 29, 2026 •

edited

Loading