1 ==============================
2 Device-mapper snapshot support
3 ==============================
5 Device-mapper allows you, without massive data copying:
7 - To create snapshots of any block device i.e. mountable, saved states of
8 the block device which are also writable without interfering with the
10 - To create device "forks", i.e. multiple different versions of the
12 - To merge a snapshot of a block device back into the snapshot's origin
15 In the first two cases, dm copies only the chunks of data that get
16 changed and uses a separate copy-on-write (COW) block device for
19 For snapshot merge the contents of the COW storage are merged back into
23 There are three dm targets available:
24 snapshot, snapshot-origin, and snapshot-merge.
26 - snapshot-origin <origin>
28 which will normally have one or more snapshots based on it.
29 Reads will be mapped directly to the backing device. For each write, the
30 original data will be saved in the <COW device> of each snapshot to keep
31 its visible content unchanged, at least until the <COW device> fills up.
34 - snapshot <origin> <COW device> <persistent?> <chunksize>
35 [<# feature args> [<arg>]*]
37 A snapshot of the <origin> block device is created. Changed chunks of
38 <chunksize> sectors will be stored on the <COW device>. Writes will
39 only go to the <COW device>. Reads will come from the <COW device> or
40 from <origin> for unchanged data. <COW device> will often be
41 smaller than the origin and if it fills up the snapshot will become
42 useless and be disabled, returning errors. So it is important to monitor
43 the amount of free space and expand the <COW device> before it fills up.
45 <persistent?> is P (Persistent) or N (Not persistent - will not survive
46 after reboot). O (Overflow) can be added as a persistent store option
47 to allow userspace to advertise its support for seeing "Overflow" in the
48 snapshot status. So supported store types are "P", "PO" and "N".
50 The difference between persistent and transient is with transient
51 snapshots less metadata must be saved on disk - they can be kept in
54 When loading or unloading the snapshot target, the corresponding
55 snapshot-origin or snapshot-merge target must be suspended. A failure to
56 suspend the origin target could result in data corruption.
60 discard_zeroes_cow - a discard issued to the snapshot device that
61 maps to entire chunks to will zero the corresponding exception(s) in
62 the snapshot's exception store.
64 discard_passdown_origin - a discard to the snapshot device is passed
65 down to the snapshot-origin's underlying device. This doesn't cause
66 copy-out to the snapshot exception store because the snapshot-origin
69 The discard_passdown_origin feature depends on the discard_zeroes_cow
70 feature being enabled.
73 - snapshot-merge <origin> <COW device> <persistent> <chunksize>
74 [<# feature args> [<arg>]*]
76 takes the same table arguments as the snapshot target except it only
77 works with persistent snapshots. This target assumes the role of the
78 "snapshot-origin" target and must not be loaded if the "snapshot-origin"
79 is still present for <origin>.
81 Creates a merging snapshot that takes control of the changed chunks
82 stored in the <COW device> of an existing snapshot, through a handover
83 procedure, and merges these chunks back into the <origin>. Once merging
84 has started (in the background) the <origin> may be opened and the merge
85 will continue while I/O is flowing to it. Changes to the <origin> are
86 deferred until the merging snapshot's corresponding chunk(s) have been
87 merged. Once merging has started the snapshot device, associated with
88 the "snapshot" target, will return -EIO when accessed.
91 How snapshot is used by LVM2
92 ============================
93 When you create the first LVM2 snapshot of a volume, four dm devices are used:
95 1) a device containing the original mapping table of the source volume;
96 2) a device used as the <COW device>;
97 3) a "snapshot" device, combining #1 and #2, which is the visible snapshot
99 4) the "original" volume (which uses the device number used by the original
100 source volume), whose table is replaced by a "snapshot-origin" mapping
103 A fixed naming scheme is used, so with the following commands::
105 lvcreate -L 1G -n base volumeGroup
106 lvcreate -L 100M --snapshot -n snap volumeGroup/base
108 we'll have this situation (with volumes in above order)::
110 # dmsetup table|grep volumeGroup
112 volumeGroup-base-real: 0 2097152 linear 8:19 384
113 volumeGroup-snap-cow: 0 204800 linear 8:19 2097536
114 volumeGroup-snap: 0 2097152 snapshot 254:11 254:12 P 16
115 volumeGroup-base: 0 2097152 snapshot-origin 254:11
117 # ls -lL /dev/mapper/volumeGroup-*
118 brw------- 1 root root 254, 11 29 ago 18:15 /dev/mapper/volumeGroup-base-real
119 brw------- 1 root root 254, 12 29 ago 18:15 /dev/mapper/volumeGroup-snap-cow
120 brw------- 1 root root 254, 13 29 ago 18:15 /dev/mapper/volumeGroup-snap
121 brw------- 1 root root 254, 10 29 ago 18:14 /dev/mapper/volumeGroup-base
124 How snapshot-merge is used by LVM2
125 ==================================
126 A merging snapshot assumes the role of the "snapshot-origin" while
127 merging. As such the "snapshot-origin" is replaced with
128 "snapshot-merge". The "-real" device is not changed and the "-cow"
129 device is renamed to <origin name>-cow to aid LVM2's cleanup of the
130 merging snapshot after it completes. The "snapshot" that hands over its
131 COW device to the "snapshot-merge" is deactivated (unless using lvchange
132 --refresh); but if it is left active it will simply return I/O errors.
134 A snapshot will merge into its origin with the following command::
136 lvconvert --merge volumeGroup/snap
138 we'll now have this situation::
140 # dmsetup table|grep volumeGroup
142 volumeGroup-base-real: 0 2097152 linear 8:19 384
143 volumeGroup-base-cow: 0 204800 linear 8:19 2097536
144 volumeGroup-base: 0 2097152 snapshot-merge 254:11 254:12 P 16
146 # ls -lL /dev/mapper/volumeGroup-*
147 brw------- 1 root root 254, 11 29 ago 18:15 /dev/mapper/volumeGroup-base-real
148 brw------- 1 root root 254, 12 29 ago 18:16 /dev/mapper/volumeGroup-base-cow
149 brw------- 1 root root 254, 10 29 ago 18:16 /dev/mapper/volumeGroup-base
152 How to determine when a merging is complete
153 ===========================================
154 The snapshot-merge and snapshot status lines end with:
156 <sectors_allocated>/<total_sectors> <metadata_sectors>
158 Both <sectors_allocated> and <total_sectors> include both data and metadata.
159 During merging, the number of sectors allocated gets smaller and
160 smaller. Merging has finished when the number of sectors holding data
161 is zero, in other words <sectors_allocated> == <metadata_sectors>.
163 Here is a practical example (using a hybrid of lvm and dmsetup commands)::
166 LV VG Attr LSize Origin Snap% Move Log Copy% Convert
167 base volumeGroup owi-a- 4.00g
168 snap volumeGroup swi-a- 1.00g base 18.97
170 # dmsetup status volumeGroup-snap
171 0 8388608 snapshot 397896/2097152 1560
172 ^^^^ metadata sectors
174 # lvconvert --merge -b volumeGroup/snap
175 Merging of volume snap started.
177 # lvs volumeGroup/snap
178 LV VG Attr LSize Origin Snap% Move Log Copy% Convert
179 base volumeGroup Owi-a- 4.00g 17.23
181 # dmsetup status volumeGroup-base
182 0 8388608 snapshot-merge 281688/2097152 1104
184 # dmsetup status volumeGroup-base
185 0 8388608 snapshot-merge 180480/2097152 712
187 # dmsetup status volumeGroup-base
188 0 8388608 snapshot-merge 16/2097152 16
190 Merging has finished.
195 LV VG Attr LSize Origin Snap% Move Log Copy% Convert
196 base volumeGroup owi-a- 4.00g