missconfigured PCI deployment
|Category:||Core & System|
|Target version:||Release 5.0.1|
|Affected Versions:||OpenNebula 5.0|
I recently upgraded from 4.14.4 to 5.0. (Looks awesome! Well done!)
I am now having an issue that I am not able to boot VMs with PCI device attached. I found out the file /var/lib/one/datastores/0/120/deployment.$i is malformed.
root@one:~ # diff wrong.pci.deployment correct.pci.deployment
< <address domain='0x0000'bus='0x00'slot='0x1d'function='0x0'/>
<address domain='0x0000' bus='0x00' slot='0x1d' function='0x0'/>
You might see the difference. After fixing the malformed I still was not able to boot the VM. After removing the complete PCI part I finally was able to boot the VM via manually executing virsh (not throught sunstone).
(This is my first ticket - please give me short notice if there are informations missing)
#1 Updated by Ruben S. Montero about 5 years ago
- Category set to Core & System
- Status changed from Pending to Closed
- Target version set to Release 5.2
- Resolution set to fixed
THANKS for the feedbak. Fix has been commited upstream.
This will address the malformed address element in the deployment file. About the VM not able to boot after fixing it, I think it might be related to your local configuration. Maybe you can get more information in the qemu/libvirt logs.
I'll close the issue, and will reopen if the qemu/libivrt problem is something related to OpenNebula