How to manage storage volumes¶
See the following sections for instructions on how to create, configure, view and resize Storage volumes.
Create a custom storage volume¶
When you create an instance, LXD automatically creates a storage volume that is used as the root disk for the instance.
You can add custom storage volumes to your instances.
Such custom storage volumes are independent of the instance, which means that they can be backed up separately and are retained until you delete them.
Custom storage volumes with content type
filesystem can also be shared between different instances.
See Storage volumes for detailed information.
Create the volume¶
Use the following command to create a custom storage volume of type
filesystem in a storage pool:
lxc storage volume create <pool_name> <volume_name> [configuration_options...]
See the Storage drivers documentation for a list of available storage volume configuration options for each driver.
By default, custom storage volumes use the
filesystem content type.
To create a custom storage volume with the content type
block, add the
lxc storage volume create <pool_name> <volume_name> --type=block [configuration_options...]
To add a custom storage volume on a cluster member, add the
lxc storage volume create <pool_name> <volume_name> --target=<cluster_member> [configuration_options...]
For most storage drivers, custom storage volumes are not replicated across the cluster and exist only on the member for which they were created.
This behavior is different for Ceph-based storage pools (
cephfs), where volumes are available from any cluster member.
To create a custom storage volume of type
iso, use the
import command instead of the
lxc storage volume import <pool_name> <iso_path> <volume_name> --type=iso
Attach the volume to an instance¶
After creating a custom storage volume, you can add it to one or more instances as a disk device.
The following restrictions apply:
Custom storage volumes of content type
isocannot be attached to containers, but only to virtual machines.
To avoid data corruption, storage volumes of content type
blockshould never be attached to more than one virtual machine at a time.
Storage volumes of content type
isoare always read-only, and can therefore be attached to more than one virtual machine at a time without corrupting data.
File system storage volumes can’t be attached to virtual machines while they’re running.
For custom storage volumes with the content type
filesystem, use the following command, where
<location> is the path for accessing the storage volume inside the instance (for example,
lxc storage volume attach <pool_name> <filesystem_volume_name> <instance_name> <location>
Custom storage volumes with the content type
block do not take a location:
lxc storage volume attach <pool_name> <block_volume_name> <instance_name>
By default, the custom storage volume is added to the instance with the volume name as the device name. If you want to use a different device name, you can add it to the command:
lxc storage volume attach <pool_name> <filesystem_volume_name> <instance_name> <device_name> <location> lxc storage volume attach <pool_name> <block_volume_name> <instance_name> <device_name>
Attach the volume as a device¶
lxc storage volume attach command is a shortcut for adding a disk device to an instance.
Alternatively, you can add a disk device for the storage volume in the usual way:
lxc config device add <instance_name> <device_name> disk pool=<pool_name> source=<volume_name> [path=<location>]
When using this way, you can add further configuration to the command if needed. See disk device for all available device options.
Configure I/O limits¶
When you attach a storage volume to an instance as a disk device, you can configure I/O limits for it.
To do so, set the
limits.max properties to the corresponding limits.
See the Type: disk reference for more information.
The limits are applied through the Linux
blkio cgroup controller, which makes it possible to restrict I/O at the disk level (but nothing finer grained than that).
Because the limits apply to a whole physical disk rather than a partition or path, the following restrictions apply:
Limits will not apply to file systems that are backed by virtual devices (for example, device mapper).
If a file system is backed by multiple block devices, each device will get the same limit.
If two disk devices that are backed by the same disk are attached to the same instance, the limits of the two devices will be averaged.
All I/O limits only apply to actual block device access. Therefore, consider the file system’s own overhead when setting limits. Access to cached data is not affected by the limit.
Use the volume for backups or images¶
To do so, you must set the corresponding server configuration:
To use a custom volume to store the backup tarballs:
lxc config set storage.backups_volume <pool_name>/<volume_name>
To use a custom volume to store the image tarballs:
lxc config set storage.images_volume <pool_name>/<volume_name>
Configure storage volume settings¶
See the Storage drivers documentation for the available configuration options for each storage driver.
Use the following command to set configuration options for a storage volume:
lxc storage volume set <pool_name> [<volume_type>/]<volume_name> <key> <value>
The default storage volume type is
custom, so you can leave out the
<volume_type>/ when configuring a custom storage volume.
For example, to set the size of your custom storage volume
my-volume to 1 GiB, use the following command:
lxc storage volume set my-pool my-volume size=1GiB
To set the snapshot expiry time for your virtual machine
my-vm to one month, use the following command:
lxc storage volume set my-pool virtual-machine/my-vm snapshots.expiry 1M
You can also edit the storage volume configuration by using the following command:
lxc storage volume edit <pool_name> [<volume_type>/]<volume_name>
Configure default values for storage volumes¶
You can define default volume configurations for a storage pool.
To do so, set a storage pool configuration with a
volume prefix, thus
This value is then used for all new storage volumes in the pool, unless it is set explicitly for a volume or an instance.
In general, the defaults set on a storage pool level (before the volume was created) can be overridden through the volume configuration, and the volume configuration can be overridden through the instance configuration (for storage volumes of type
For example, to set a default volume size for a storage pool, use the following command:
lxc storage set [<remote>:]<pool_name> volume.size <value>
View storage volumes¶
You can display a list of all available storage volumes in a storage pool and check their configuration.
To list all available storage volumes in a storage pool, use the following command:
lxc storage volume list <pool_name>
To display the storage volumes for all projects (not only the default project), add the
Custom storage volumes might use the same name as instance volumes (for example, you might have a container named
c1 with a container storage volume named
c1 and a custom storage volume named
Therefore, to distinguish between instance storage volumes and custom storage volumes, all instance storage volumes must be referred to as
<volume_type>/<volume_name> (for example,
virtual-machine/vm) in commands.
To show detailed configuration information about a specific volume, use the following command:
lxc storage volume show <pool_name> [<volume_type>/]<volume_name>
To show state information about a specific volume, use the following command:
lxc storage volume info <pool_name> [<volume_type>/]<volume_name>
In both commands, the default storage volume type is
custom, so you can leave out the
<volume_type>/ when displaying information about a custom storage volume.
Resize a storage volume¶
If you need more storage in a volume, you can increase the size of your storage volume. In some cases, it is also possible to reduce the size of a storage volume.
To resize a storage volume, set its size configuration:
lxc storage volume set <pool_name> <volume_name> size <new_size>
Growing a storage volume usually works (if the storage pool has sufficient storage).
Shrinking a storage volume is only possible for storage volumes with content type
filesystem. It is not guaranteed to work though, because you cannot shrink storage below its current used size.
Shrinking a storage volume with content type
blockis not possible.