3
A qcow2 image file is organized in units of constant size, which are called
4
(host) clusters. A cluster is the unit in which all allocations are done,
5
both for actual guest data and for image metadata.
7
Likewise, the virtual disk as seen by the guest is divided into (guest)
8
clusters of the same size.
10
All numbers in qcow2 are stored in Big Endian byte order.
15
The first cluster of a qcow2 image contains the file header:
18
QCOW magic string ("QFI\xfb")
21
Version number (only valid value is 2)
23
8 - 15: backing_file_offset
24
Offset into the image file at which the backing file name
25
is stored (NB: The string is not null terminated). 0 if the
26
image doesn't have a backing file.
28
16 - 19: backing_file_size
29
Length of the backing file name in bytes. Must not be
30
longer than 1023 bytes. Undefined if the image doesn't have
34
Number of bits that are used for addressing an offset
35
within a cluster (1 << cluster_bits is the cluster size).
36
Must not be less than 9 (i.e. 512 byte clusters).
38
Note: qemu as of today has an implementation limit of 2 MB
39
as the maximum cluster size and won't be able to open images
40
with larger cluster sizes.
43
Virtual disk size in bytes
50
Number of entries in the active L1 table
52
40 - 47: l1_table_offset
53
Offset into the image file at which the active L1 table
54
starts. Must be aligned to a cluster boundary.
56
48 - 55: refcount_table_offset
57
Offset into the image file at which the refcount table
58
starts. Must be aligned to a cluster boundary.
60
56 - 59: refcount_table_clusters
61
Number of clusters that the refcount table occupies
64
Number of snapshots contained in the image
66
64 - 71: snapshots_offset
67
Offset into the image file at which the snapshot table
68
starts. Must be aligned to a cluster boundary.
70
Directly after the image header, optional sections called header extensions can
71
be stored. Each extension has a structure like the following:
73
Byte 0 - 3: Header extension type:
74
0x00000000 - End of the header extension area
75
0xE2792ACA - Backing file format name
76
other - Unknown header extension, can be safely
79
4 - 7: Length of the header extension data
81
8 - n: Header extension data
83
n - m: Padding to round up the header extension size to the next
86
The remaining space between the end of the header extension area and the end of
87
the first cluster can be used for other data. Usually, the backing file name is
91
== Host cluster management ==
93
qcow2 manages the allocation of host clusters by maintaining a reference count
94
for each host cluster. A refcount of 0 means that the cluster is free, 1 means
95
that it is used, and >= 2 means that it is used and any write access must
96
perform a COW (copy on write) operation.
98
The refcounts are managed in a two-level table. The first level is called
99
refcount table and has a variable size (which is stored in the header). The
100
refcount table can cover multiple clusters, however it needs to be contiguous
103
It contains pointers to the second level structures which are called refcount
104
blocks and are exactly one cluster in size.
106
Given a offset into the image file, the refcount of its cluster can be obtained
109
refcount_block_entries = (cluster_size / sizeof(uint16_t))
111
refcount_block_index = (offset / cluster_size) % refcount_table_entries
112
refcount_table_index = (offset / cluster_size) / refcount_table_entries
114
refcount_block = load_cluster(refcount_table[refcount_table_index]);
115
return refcount_block[refcount_block_index];
117
Refcount table entry:
119
Bit 0 - 8: Reserved (set to 0)
121
9 - 63: Bits 9-63 of the offset into the image file at which the
122
refcount block starts. Must be aligned to a cluster
125
If this is 0, the corresponding refcount block has not yet
126
been allocated. All refcounts managed by this refcount block
129
Refcount block entry:
131
Bit 0 - 15: Reference count of the cluster
134
== Cluster mapping ==
136
Just as for refcounts, qcow2 uses a two-level structure for the mapping of
137
guest clusters to host clusters. They are called L1 and L2 table.
139
The L1 table has a variable size (stored in the header) and may use multiple
140
clusters, however it must be contiguous in the image file. L2 tables are
141
exactly one cluster in size.
143
Given a offset into the virtual disk, the offset into the image file can be
146
l2_entries = (cluster_size / sizeof(uint64_t))
148
l2_index = (offset / cluster_size) % l2_entries
149
l1_index = (offset / cluster_size) / l2_entries
151
l2_table = load_cluster(l1_table[l1_index]);
152
cluster_offset = l2_table[l2_index];
154
return cluster_offset + (offset % cluster_size)
158
Bit 0 - 8: Reserved (set to 0)
160
9 - 55: Bits 9-55 of the offset into the image file at which the L2
161
table starts. Must be aligned to a cluster boundary. If the
162
offset is 0, the L2 table and all clusters described by this
163
L2 table are unallocated.
165
56 - 62: Reserved (set to 0)
167
63: 0 for an L2 table that is unused or requires COW, 1 if its
168
refcount is exactly one. This information is only accurate
169
in the active L1 table.
171
L2 table entry (for normal clusters):
173
Bit 0 - 8: Reserved (set to 0)
175
9 - 55: Bits 9-55 of host cluster offset. Must be aligned to a
176
cluster boundary. If the offset is 0, the cluster is
179
56 - 61: Reserved (set to 0)
181
62: 0 (this cluster is not compressed)
183
63: 0 for a cluster that is unused or requires COW, 1 if its
184
refcount is exactly one. This information is only accurate
185
in L2 tables that are reachable from the the active L1
188
L2 table entry (for compressed clusters; x = 62 - (cluster_size - 8)):
190
Bit 0 - x: Host cluster offset. This is usually _not_ aligned to a
193
x+1 - 61: Compressed size of the images in sectors of 512 bytes
195
62: 1 (this cluster is compressed using zlib)
197
63: 0 for a cluster that is unused or requires COW, 1 if its
198
refcount is exactly one. This information is only accurate
199
in L2 tables that are reachable from the the active L1
202
If a cluster is unallocated, read requests shall read the data from the backing
203
file. If there is no backing file or the backing file is smaller than the image,
204
they shall read zeros for all parts that are not covered by the backing file.
209
qcow2 supports internal snapshots. Their basic principle of operation is to
210
switch the active L1 table, so that a different set of host clusters are
211
exposed to the guest.
213
When creating a snapshot, the L1 table should be copied and the refcount of all
214
L2 tables and clusters reachable form this L1 table must be increased, so that
215
a write causes a COW and isn't visible in other snapshots.
217
When loading a snapshot, bit 63 of all entries in the new active L1 table and
218
all L2 tables referenced by it must be reconstructed from the refcount table
219
as it doesn't need to be accurate in inactive L1 tables.
221
A directory of all snapshots is stored in the snapshot table, a contiguous area
222
in the image file, whose starting offset and length are given by the header
223
fields snapshots_offset and nb_snapshots. The entries of the snapshot table
224
have variable length, depending on the length of ID, name and extra data.
226
Snapshot table entry:
228
Byte 0 - 7: Offset into the image file at which the L1 table for the
229
snapshot starts. Must be aligned to a cluster boundary.
231
8 - 11: Number of entries in the L1 table of the snapshots
233
12 - 13: Length of the unique ID string describing the snapshot
235
14 - 15: Length of the name of the snapshot
237
16 - 19: Time at which the snapshot was taken in seconds since the
240
20 - 23: Subsecond part of the time at which the snapshot was taken
243
24 - 31: Time that the guest was running until the snapshot was
246
32 - 35: Size of the VM state in bytes. 0 if no VM state is saved.
247
If there is VM state, it starts at the first cluster
248
described by first L1 table entry that doesn't describe a
249
regular guest cluster (i.e. VM state is stored like guest
250
disk content, except that it is stored at offsets that are
251
larger than the virtual disk presented to the guest)
253
36 - 39: Size of extra data in the table entry (used for future
254
extensions of the format)
256
variable: Extra data for future extensions. Must be ignored.
258
variable: Unique ID string for the snapshot (not null terminated)
260
variable: Name of the snapshot (not null terminated)