English German
Creating a Logical Volume Managament (LVM) Layout
_Logical Volume Management_ (LVM) presents a simple logical view of underlying physical storage space, such as hard drives or LUNs. Partitions on physical storage are represented as _physical volumes_ that can be grouped together into _volume groups_. Each volume group can be divided into multiple _logical volumes_, each of which is analogous to a standard disk partition. Therefore, LVM logical volumes function as partitions which can span multiple physical disks.
See xref:appendixes/Understanding_LVM.adoc#appe-lvm-overview[Understanding LVM] for additional information about the concepts behind Logical Volume Management.
Some partition types - notably the `/boot` directory and the BIOS Boot and EFI partitions - can not be placed on logical volumes. Use standard physical volumes for them. See xref:Installing_Using_Anaconda.adoc#sect-installation-gui-manual-partitioning-recommended[Recommended Partitioning Scheme] for more information.
Create LVM Logical Volume
The Manual Partitioning screen
Follow the procedure below to create LVM logical volumes and volume groups.
Creating LVM Logical Volumes and Groups
Click the `+` button at the bottom of the list showing existing mount points. A new dialog window will open.
In the new dialog window, specify a mount point for which you want to create a separate logical volume - for example, `/`. Optionally, specify a size for the volume using standard units such as MB or GB (for example, `50GB`). Then, click `Add mount point` to add the volume and return to the main partitioning screen.
When creating a mount point for swap on LVM, specify the mount point as `swap`.
The mount point has now been created using the default settings, which means it has been created as an LVM logical volume, and a volume group has been created to contain it. Select the newly created mount point in the left pane to configure it further. If you want to use thin provisioning for this volume, change the `Device Type` option to `LVM Thin Provisioning`.
In the `Volume Group` menu, you can see that the volume has been assigned to an automatically created volume group, which is named after the {PRODUCT} variant you are installing (for example, `fedora-server`. Click the `Modify` button under the drop-down menu to access the volume group settings.
In the `Configure Volume Group` dialog, you can change the volume group's name, its `RAID level` (see xref:Installing_Using_Anaconda.adoc#sect-installation-gui-manual-partitioning-filesystems[Device, File System and RAID Types] for information about available RAID types), and you can also specify which physical devices (disks) this volume group should reside on. You can select one or more disks which will be used to hold this volume group by holding down kbd:[Ctrl] and clicking each disk in the list.
If you select a redundant RAID type (such as `RAID1 (Redundancy)`), the volume group will take up twice its actual size on your disks. A 5 GB volume group with RAID1 will take up 10 GB of space.
You can also make sure that the volume group is encrypted by selecting the `Encrypt` option; this will enable LUKS encryption for the entire volume group. See the [citetitle]_{PRODUCT} Security Guide_, available at link:++https://docs.fedoraproject.org/++[], for information about LUKS disk encryption.
Additionally, you can set a fixed size for the volume group by selecting the `Fixed` option from the `Size policy` menu and entering a size for the volume group.
After you finish configuring the volume group settings, click `Save` to return to the main `Manual Partitioning` screen.
The configuration dialog does not allow you to specify the size of the volume group's _physical extents_. The size will always be set to the default value of 4 MiB. If you want to create a volume group with different physical extents, create it manually by switching to an interactive shell and using the [command]#vgcreate# command, or use a Kickstart file with the [command]#volgroup --pesize=pass:attributes[{blank}]_size_pass:attributes[{blank}]# command.