Cannot boot VM with more than one scsi disk

Description Booting a VM with more than one scsi disk stops with an error message. The error occurs with adapter “lsilogic” and “buslogic”. VMs with one scsi disk work. VMs with multiple virtio disks work as well. We need scsi disks for trim support.

/var/log/one/oned.log:

LOG I 21 error: unsupported configuration: target must be 0 for controller model 'lsilogic'

Error shown in Sunstone:

Error deploying virtual machine: Could not create domain from /var/lib/one//datastores/0/21/deployment.20

Extract of /var/lib/one/datastores/0/21/deployment.20:

		<disk type='file' device='cdrom'>
			<source file='/var/lib/one//datastores/0/21/disk.0'/>
			<target dev='sda'/>
			<boot order='2'/>
			<readonly/>
			<driver name='qemu' type='raw' cache='none'/>
			<address type='drive' controller='0' bus='0' target='0' unit='0'/>
		</disk>
		<disk type='file' device='disk'>
			<source file='/var/lib/one//datastores/0/21/disk.4'/>
			<target dev='sdb'/>
			<driver name='qemu' type='qcow2' cache='none'/>
			<address type='drive' controller='0' bus='0' target='1' unit='0'/>
		</disk>

To Reproduce Create a VM, add two scsi disks, boot. We have the machine type set to “q35” globally.

Expected behavior VM should boot.

Details

• Affected Component: Scheduler or Storage
• Hypervisor: KVM
• Version: 5.6.1-1
• Machine: q35
• Architecure: x86_64

Additional context We think, the problem was introduced here: https://github.com/OpenNebula/one/commit/e164fde85f4459add2e8007ebcf15da5d045d514

See also: https://github.com/libvirt/libvirt/blob/master/src/qemu/qemu_command.c#L5103

Progress Status

• Branch created
• Code committed to development branch
• Testing - QA
• Documentation
• Release notes - resolved issues, compatibility, known issues
• Code committed to upstream release/hotfix branches
• Documentation committed to upstream release/hotfix branches

[sergiojvg]

We have reproduced the issue.

[k51312]

hi， i have a trouble in create vm with scsi disk, for the guest os can’t start

[root@NSG ~]# virsh list --all Id Name State

2 root-vsys_scsi-vm1 running

[root@NSG ~]# virsh dumpxml root-vsys_scsi-vm1

then the vm can success to create ,but when i enter into the guest os ,found the guest fail to start

my centos7 kernel: linux-3.10.0-693.17.1.el7.x86_64 ,this kernel has reduced lots of parameters. this mybe my host kernel .config has some problem, how should i config my kernel .config ??

            thanks 

scsi-test.log