Taking an LVM logical volume offline for maintenance requires more than hiding a mount point. The kernel must release the device-mapper node before storage repair, snapshot merge, removal, or handoff work can proceed, and deactivating the wrong LV can interrupt applications that still hold it open.
The lvchange command controls the activation state for one logical volume. Using --activate n makes the LV unavailable to the kernel, while --activate y restores the block device so filesystems or applications can use it again.
Activation does not mount a filesystem or restart an application. Confirm the target path first, stop or unmount anything using it, and follow the cluster resource policy before changing shared-storage LVM volumes.
Related: How to create an LVM logical volume
Related: How to remove an LVM logical volume
$ sudo lvs -o lv_name,vg_name,lv_path LV VG Path data vgdata /dev/vgdata/data
$ findmnt -o SOURCE,TARGET /dev/vgdata/data SOURCE TARGET /dev/vgdata/data /srv/data
No output means findmnt did not see that LV as a mounted filesystem. Raw consumers such as databases, virtual machines, swap, or cluster resources can still hold the device open.
$ sudo umount /srv/data
Do not deactivate a mounted or in-use LV. Stop the application, unmount the filesystem, or fail over the resource first to avoid data loss or service interruption.
$ sudo lvchange --activate n vgdata/data
$ sudo lvdisplay vgdata/data --- Logical volume --- LV Path /dev/vgdata/data LV Name data VG Name vgdata LV Status NOT available ##### snipped #####
$ sudo lvchange --activate y vgdata/data
$ sudo lvdisplay vgdata/data --- Logical volume --- LV Path /dev/vgdata/data LV Name data VG Name vgdata LV Status available ##### snipped #####
$ sudo mount /srv/data
Use the mount point from findmnt or /etc/fstab. Skip this step for raw logical volumes that are opened directly by an application.