Known Issues

A complete list of known issues for OpenNebula is maintained here.

This page will be updated with relevant information about bugs affecting OpenNebula, as well as possible workarounds until a patch is officially published.

Drivers - Network

  • If the nic-attach fails due to the libvirt bug (VM can not eject CD-ROM after reset) the nic appears in the VM (although without proper configuration) but it’s not visible on OpenNebula VM #5268

  • Edge Cluster Public IP: NIC_ALIAS on the public network can only can only be associated to a NIC on the same network.

Drivers - Storage

  • LXC, XFS formatted disk images are incompatible with the fs_lvm driver. The image fails to be mounted on the host.

High Availability

HA server configuration synchronization with the command onezone serversync does not work for the PostgreSQL and SQLite DB Back-end

Sunstone

  • Guacamole RDP as is currently shipped in OpenNebula does not support NLA authentication. You can follow these instructions in order to disable NLA in the Windows box to use Guacamole RDP within Sunstone.

  • When hidding the VM name setting on instantiate form, Sunstone does not continue with the instantintiation process. To be able to instantiate templates please set VM.instantiate_name to true in the corresponding YAML file.

Install Linux Graphical Desktop on KVM Virtual Machines

OpenNebula uses the cirrus graphical adapter for KVM Virtual Machines by default. It could happen that after installing a graphical desktop on a Linux VM, the Xorg window system does not load the appropriate video driver. You can force a VESA mode by configuring the kernel parameter vga=VESA_MODE in the GNU GRUB configuration file. Here you can find the VESA mode numbers. For example, adding vga=791 as kernel parameter will select the 16-bit 1024×768 resolution mode.

vCenter Snapshot behavior

VMs in vCenter 7.0 exhibit a new behavior regarding snapshots and disks attach/detach operations. When vCenter 7.0 detects any change in the number of disks attached to a VM, it automatically cleans all the VM snapshots. OpenNebula doesn’t take this into account yet, so the snapshots stated by OpenNebula, after a disk attach or disk detach, are pointing to a null vCenter reference, and as such, cannot be used. Please keep this in mind before a solution is implemented.

Virtual Machines Backup

When taking a VM backup, if the upload process fails and the app results in ERROR state, the backup will complete successfully. A detailed explanation can be found here.

Warning when Exporting an App from the Marketplace Using CLI

When exporting an application from the marketplace using the CLI the following warning can be seen:

/usr/lib/one/ruby/opennebula/xml_element.rb:124: warning: Passing a Node as the second parameter to Node.new is deprecated. Please pass a Document instead, or prefer an alternative constructor like Node#add_child. This will become an error in a future release of Nokogiri.

This is harmless and can be discarded, it will be addressed in future releases.

Contextualization

  • GROW_ROOTFS and GROW_FS will not extend btrfs filesystems

  • Debian 12 (pre-release) has issue with systemd-networkd-wait-online which timeouts and delays contextualization

FireEdge

  • When creating an onprem provision the second step (Provider) shows and empty list instead of the onprem provider. In order to workaround the issue just click on the next button at the top right corner.

  • When deploying an HCI onprem provision it will fails because of the Ansible version used. In order to workaround the issue you need to fix the value of CEPH_ANSIBLE_BRANCH to v6.0.26 at /usr/lib/one/oneprovision/lib/provision/ansible.rb file in the Front-end node.

OneProvision Equinix Provider

Due to changes in the facilities names in Equinix, you need to manually update the provider files share/oneprovision/edge-clusters/metal/providers/equinix. As a reference, you can use the following change to update your files.

OneFlow

Some services may stuck in transient states (DEPLOYING, SCALING or UNDEPLOYING), this is due to expiration of the serveradmin token. There is already an issue opened to fix it.

KVM - Live Memory Resize

For live memory resize user needs to specify MAX_MEMORY, the available memory for the VM is in MEMORY attribute. But VM running Windows shows in Task Manager available physical memory as MAX_MEMORY although it can use only MEMORY, this value is not visible in the Task Manager. Because of this it shows a high percentage of memory used. To double check the real available memory for the VM user should use command virsh dominfo <id> on the host.