LVM snapshots are logical volumes that reflect the state of the snapshoted volume at the exact moment in time the snapshot was created. Useful for backups and reference points we can revert back to.
$ sudo lvcreate --size 5G --snapshot --name root-backup /dev/vg0/root
Logical volume "root-backup" created.We created a new volume named root-backup, of 5GB size and type snapshot. The target volume for which we created a snapshot is /dev/vg0/root (my root partition).
The 5 GB size is a buffer area. Until the changes made on our root partition fit inside that preallocated space, we are able to revert back to the point referenced by the snapshot. When we pass that threshold the snapshot will become inactive (= useless).
With lvs we can check the available space left on the snapshot buffer. (Or with lvdisplay, but that would be more verbose for our usecase)
$ sudo lvs
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
docker-pool vg0 twi-a-t--- 3.80g 0.00 0.02
root vg0 owi-aos--- 238.00g
root-backup vg0 swi-a-s--- 5.00g root 0.00Next I install all the packages that start with the gnome prefix, and see how that changes usage.
$ sudo dnf install "gnome-*"
[... ignored lines ...]
Install 589 Packages
Total download size: 519 M
Installed size: 1.4 G
$ sudo lvs
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
docker-pool vg0 twi-a-t--- 3.80g 0.00 0.02
root vg0 owi-aos--- 238.00g
root-backup vg0 swi-a-s--- 5.00g root 47.56$ sudo lvconvert --merge /dev/vg0/root-backup
Logical volume vg0/root contains a filesystem in use.
Can't merge over open origin volume.
Merging of snapshot vg0/root-backup will occur on next activation of vg0/root.As seen from the message the restoration will happen at the next activation of vg0/root. In my case that means a system reboot. You might want to drop down to a runlevel 3, or multi-user.target / rescue.target (since everybody's using systemd nowadays), and unmount and restore that way. Reboot looks simpler for me.
In order to persist the changes that have been made since the snapshot is created you can just remove the snapshot logical volume.
$ sudo lvremove /dev/vg0/root-backup
Do you really want to remove active logical volume root-backup? [y/n]: y
Logical volume "root-backup" successfully removed- Unless you've allocated space in advance in your volume group for snapshots you won't be able to use this functionality. On a default install you might allocate all the disk space to a single logical volume like I did. In that scenario you can create somme free space for snapshoting by downsizing your logical volumes. I wasn't forced to do that since I had another hard drive on which I've created a new LVM physical volume, and attached that to my main volume group.
To check if you have enough space to create snapshots, use the vgdisplay command and check the "Free PE / Size" field
$ sudo vgdisplay
--- Volume group ---
VG Name vg0
System ID
Format lvm2
Metadata Areas 2
Metadata Sequence No 66
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 2
Open LV 1
Max PV 0
Cur PV 2
Act PV 2
VG Size 247.76 GiB
PE Size 4.00 MiB
Total PE 63427
Alloc PE / Size 62030 / 242.30 GiB
Free PE / Size 1397 / 5.46 GiB
VG UUID 819Sbr-dcnl-45Ym-7ety-b5TH-MNqZ-2L5rGM- The merge isn't instant and for my example it took a couple of minutes for the merge to complete. You can see if a merge is still in progress using lvdisplay and checking if the "LV Snapshot Status" field is present. When the merging is done that field won't show up and vgdisplay will show the snapshot allocated size as free space once again
sudo lvdisplay
--- Logical volume ---
LV Path /dev/vg0/root
LV Name root
VG Name vg0
LV UUID b3bnyS-O4Ju-4vQx-13LK-zEJq-gvE6-N4GkbX
LV Write Access read/write
LV Creation host, time x, 2016-02-17 00:23:24 +0200
LV snapshot status source of
root-backup [active]
LV Status available
# open 1
LV Size 238.00 GiB
Current LE 60928
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:2