Hide Forgot
Description of problem: Make sure there existed an guest that could be started. Remove its hard disk,(assume its disk type is "ide") then add an "scsi" disk; Remove the "scsi" disk, and add its original "ide" hard disk, the guest could not started with an error. Version-Release number of selected component (if applicable): libvirt-0.8.2-18.el5 virt-manager-0.6.1-14.el5 kernel-2.6.18-259.el5 How reproducible: always Steps to Reproduce: Make sure there existed an guest that could be started.(assume its disk type is "ide") 1.Edit->Virtual Machine Details->Hardware, Remove the "Disk hda" 2.Add Hardware->Storage, select "scsi" disk image, finish 3.Remove the new added "scsi" Disk 4.Add Hardware->Storage, select the original "ide" disk 5.The guest could not be started with an error. Actual results: "Error starting domain: internal error cannot find PCI address for controller scsi" Expected results: The guest could be started successfully. Additional info: #virsh dumpxml $guest <disk type='file' device='disk'> <driver name='qemu' type='raw'/> <source file='/var/lib/libvirt/images/4.9.img'/> <target dev='hda' bus='ide'/> <address type='drive' controller='0' bus='0' unit='0'/> </disk> <controller type='ide' index='0'/> <controller type='scsi' index='0'/>
This request was evaluated by Red Hat Product Management for inclusion in Red Hat Enterprise Linux 5.7 and Red Hat does not plan to fix this issue the currently developed update. Contact your manager or support representative in case you need to escalate this bug.
I think adding an option in the UI to remove the SCSI controller is too invasive of a change at this point. There is already another bug about hiding the SCSI option which should prevent people from getting into this situation. I'm okay with recommending that people who have already got a messed up config need to go to the command line to fix it. Closing this as WONTFIX for RHEL5