The LVM datastore driver provides OpenNebula with the possibility of using LVM volumes instead of plain files to hold the Virtual Images. This reduces the overhead of having a file-system in place and thus it may increase I/O performance.
Images are stored as regular files (under the usual path:
/var/lib/one/datastores/<id>) in the Image Datastore, but they will be dumped into a Logical Volumes (LV) upon virtual machine creation. The virtual machines will run from Logical Volumes in the node.
This is the recommended driver to be used when a high-end SAN is available. The same LUN can be exported to all the hosts, Virtual Machines will be able to run directly from the SAN.
The LVM datastore does not need CLVM configured in your cluster. The drivers refresh LVM meta-data each time an image is needed in another host.
For example, consider a system with two Virtual Machines (9 and 10) using a disk, running in a LVM Datastore, with ID 0. The nodes have configured a shared LUN and created a volume group named
vg-one-0, the layout of the datastore would be:
lvs LV VG Attr LSize Pool Origin Data% Meta% Move lv-one-10-0 vg-one-0 -wi------- 2.20g lv-one-9-0 vg-one-0 -wi------- 2.20g
Images are stored in a shared storage in file form (e.g. NFS, GlusterFS...) the datastore directories and mount points needs to be configured as a regular shared image datastore, please refer to FileSystem Datastore guide. It is a good idea to first deploy a shared FileSystem datastore and once it is working replace the associated System datastore with the LVM one maintaining the shared mount point as described below.
- The frontend needs to have access to the images datastore, mounting the associated directory.
Nodes needs to meet the following requirements:
- LVM2 must be available in the Hosts.
lvmetadmust be disabled. Set this parameter in
use_lvmetad = 0, and disable the
oneadminneeds to belong to the
- All the nodes needs to have access to the same LUNs.
- A LVM VG needs to be created in the shared LUNs for each datastore following name:
vg-one-<system_ds_id>. This just need to be done in one node.
- Virtual Machine disks are symbolic links to the block devices. However, additional VM files like checkpoints or deployment files are stored under
/var/lib/one/datastores/<id>. Be sure that enough local space is present.
- All the nodes needs to have access to the images and system datastores, mounting the associated directories.
In case of the virtualization host reboot, the volumes need to be activated to be available for the hypervisor again. If the node package is installed, the activation is done automatically. Otherwise, each volume device of the virtual machines running on the host before the reboot needs to be activated manually by running
lvchange -ay $DEVICE (or, activation script
/var/tmp/one/tm/fs_lvm/activate from the remote scripts may be executed on the host to do the job).
Once the storage is setup, the OpenNebula configuration comprises two steps:
- Create a System Datastore
- Create an Image Datastore
Create a System Datastore¶
LVM System Datastores needs to be created with the following values:
||The name of the Datastore|
||List of nodes with access to the LV to monitor it and also used when deleting the LV.|
> cat ds.conf NAME = lvm_system TM_MAD = fs_lvm TYPE = SYSTEM_DS BRIDGE_LIST = node1.kvm.lvm node2.kvm.lvm > onedatastore create ds.conf ID: 100
Create an Image Datastore¶
To create an Image Datastore you just need to define the name, and set the following:
||The name of the datastore|
For example, the following examples illustrates the creation of an LVM datastore using a configuration file. In this case we will use the host
host01 as one of our OpenNebula LVM-enabled hosts.
> cat ds.conf NAME = production DS_MAD = fs TM_MAD = fs_lvm DISK_TYPE = "BLOCK" TYPE = IMAGE_DS SAFE_DIRS="/var/tmp /tmp" > onedatastore create ds.conf ID: 101
By default the LVM driver will zero any LVM volume so VM data cannot leak to other instances. However, this process takes some time and my delay the deployment of a VM. The behavior of the driver can be configured in the file
/var/lib/one/remotes/etc/fs_lvm/fs_lvm.conf, in particular:
||Zero LVM volumes when they are created/resized|
||Zero LVM volumes when VM disks are deleted|
# Zero LVM volumes on creation or resizing ZERO_LVM_ON_CREATE=no # Zero LVM volumes on delete, when the VM disks are disposed ZERO_LVM_ON_DELETE=yes