xref: /openbmc/qemu/docs/tools/qemu-img.rst (revision 886cc689)
1QEMU disk image utility
2=======================
3
4Synopsis
5--------
6
7**qemu-img** [*standard options*] *command* [*command options*]
8
9Description
10-----------
11
12qemu-img allows you to create, convert and modify images offline. It can handle
13all image formats supported by QEMU.
14
15**Warning:** Never use qemu-img to modify images in use by a running virtual
16machine or any other process; this may destroy the image. Also, be aware that
17querying an image that is being modified by another process may encounter
18inconsistent state.
19
20Options
21-------
22
23.. program:: qemu-img
24
25Standard options:
26
27.. option:: -h, --help
28
29  Display this help and exit
30
31.. option:: -V, --version
32
33  Display version information and exit
34
35.. option:: -T, --trace [[enable=]PATTERN][,events=FILE][,file=FILE]
36
37  .. include:: ../qemu-option-trace.rst.inc
38
39The following commands are supported:
40
41.. hxtool-doc:: qemu-img-cmds.hx
42
43Command parameters:
44
45*FILENAME* is a disk image filename.
46
47*FMT* is the disk image format. It is guessed automatically in most
48cases. See below for a description of the supported disk formats.
49
50*SIZE* is the disk image size in bytes. Optional suffixes ``k`` or
51``K`` (kilobyte, 1024) ``M`` (megabyte, 1024k) and ``G`` (gigabyte,
521024M) and T (terabyte, 1024G) are supported.  ``b`` is ignored.
53
54*OUTPUT_FILENAME* is the destination disk image filename.
55
56*OUTPUT_FMT* is the destination format.
57
58*OPTIONS* is a comma separated list of format specific options in a
59name=value format. Use ``-o ?`` for an overview of the options supported
60by the used format or see the format descriptions below for details.
61
62*SNAPSHOT_PARAM* is param used for internal snapshot, format is
63'snapshot.id=[ID],snapshot.name=[NAME]' or '[ID_OR_NAME]'.
64
65..
66  Note the use of a new 'program'; otherwise Sphinx complains about
67  the -h option appearing both in the above option list and this one.
68
69.. program:: qemu-img-common-opts
70
71.. option:: --object OBJECTDEF
72
73  is a QEMU user creatable object definition. See the :manpage:`qemu(1)`
74  manual page for a description of the object properties. The most common
75  object type is a ``secret``, which is used to supply passwords and/or
76  encryption keys.
77
78.. option:: --image-opts
79
80  Indicates that the source *FILENAME* parameter is to be interpreted as a
81  full option string, not a plain filename. This parameter is mutually
82  exclusive with the *-f* parameter.
83
84.. option:: --target-image-opts
85
86  Indicates that the OUTPUT_FILENAME parameter(s) are to be interpreted as
87  a full option string, not a plain filename. This parameter is mutually
88  exclusive with the *-O* parameters. It is currently required to also use
89  the *-n* parameter to skip image creation. This restriction may be relaxed
90  in a future release.
91
92.. option:: --force-share (-U)
93
94  If specified, ``qemu-img`` will open the image in shared mode, allowing
95  other QEMU processes to open it in write mode. For example, this can be used to
96  get the image information (with 'info' subcommand) when the image is used by a
97  running guest.  Note that this could produce inconsistent results because of
98  concurrent metadata changes, etc. This option is only allowed when opening
99  images in read-only mode.
100
101.. option:: --backing-chain
102
103  Will enumerate information about backing files in a disk image chain. Refer
104  below for further description.
105
106.. option:: -c
107
108  Indicates that target image must be compressed (qcow format only).
109
110.. option:: -h
111
112  With or without a command, shows help and lists the supported formats.
113
114.. option:: -p
115
116  Display progress bar (compare, convert and rebase commands only).
117  If the *-p* option is not used for a command that supports it, the
118  progress is reported when the process receives a ``SIGUSR1`` or
119  ``SIGINFO`` signal.
120
121.. option:: -q
122
123  Quiet mode - do not print any output (except errors). There's no progress bar
124  in case both *-q* and *-p* options are used.
125
126.. option:: -S SIZE
127
128  Indicates the consecutive number of bytes that must contain only zeros
129  for qemu-img to create a sparse image during conversion. This value is rounded
130  down to the nearest 512 bytes. You may use the common size suffixes like
131  ``k`` for kilobytes.
132
133.. option:: -t CACHE
134
135  Specifies the cache mode that should be used with the (destination) file. See
136  the documentation of the emulator's ``-drive cache=...`` option for allowed
137  values.
138
139.. option:: -T SRC_CACHE
140
141  Specifies the cache mode that should be used with the source file(s). See
142  the documentation of the emulator's ``-drive cache=...`` option for allowed
143  values.
144
145Parameters to snapshot subcommand:
146
147.. program:: qemu-img-snapshot
148
149.. option:: snapshot
150
151  Is the name of the snapshot to create, apply or delete
152
153.. option:: -a
154
155  Applies a snapshot (revert disk to saved state)
156
157.. option:: -c
158
159  Creates a snapshot
160
161.. option:: -d
162
163  Deletes a snapshot
164
165.. option:: -l
166
167  Lists all snapshots in the given image
168
169Parameters to compare subcommand:
170
171.. program:: qemu-img-compare
172
173.. option:: -f
174
175  First image format
176
177.. option:: -F
178
179  Second image format
180
181.. option:: -s
182
183  Strict mode - fail on different image size or sector allocation
184
185Parameters to convert subcommand:
186
187.. program:: qemu-img-convert
188
189.. option:: -n
190
191  Skip the creation of the target volume
192
193.. option:: -m
194
195  Number of parallel coroutines for the convert process
196
197.. option:: -W
198
199  Allow out-of-order writes to the destination. This option improves performance,
200  but is only recommended for preallocated devices like host devices or other
201  raw block devices.
202
203.. option:: -C
204
205  Try to use copy offloading to move data from source image to target. This may
206  improve performance if the data is remote, such as with NFS or iSCSI backends,
207  but will not automatically sparsify zero sectors, and may result in a fully
208  allocated target image depending on the host support for getting allocation
209  information.
210
211.. option:: --salvage
212
213  Try to ignore I/O errors when reading.  Unless in quiet mode (``-q``), errors
214  will still be printed.  Areas that cannot be read from the source will be
215  treated as containing only zeroes.
216
217.. option:: --target-is-zero
218
219  Assume that reading the destination image will always return
220  zeros. This parameter is mutually exclusive with a destination image
221  that has a backing file. It is required to also use the ``-n``
222  parameter to skip image creation.
223
224Parameters to dd subcommand:
225
226.. program:: qemu-img-dd
227
228.. option:: bs=BLOCK_SIZE
229
230  Defines the block size
231
232.. option:: count=BLOCKS
233
234  Sets the number of input blocks to copy
235
236.. option:: if=INPUT
237
238  Sets the input file
239
240.. option:: of=OUTPUT
241
242  Sets the output file
243
244.. option:: skip=BLOCKS
245
246  Sets the number of input blocks to skip
247
248Command description:
249
250.. program:: qemu-img-commands
251
252.. option:: amend [--object OBJECTDEF] [--image-opts] [-p] [-q] [-f FMT] [-t CACHE] -o OPTIONS FILENAME
253
254  Amends the image format specific *OPTIONS* for the image file
255  *FILENAME*. Not all file formats support this operation.
256
257.. option:: bench [-c COUNT] [-d DEPTH] [-f FMT] [--flush-interval=FLUSH_INTERVAL] [-i AIO] [-n] [--no-drain] [-o OFFSET] [--pattern=PATTERN] [-q] [-s BUFFER_SIZE] [-S STEP_SIZE] [-t CACHE] [-w] [-U] FILENAME
258
259  Run a simple sequential I/O benchmark on the specified image. If ``-w`` is
260  specified, a write test is performed, otherwise a read test is performed.
261
262  A total number of *COUNT* I/O requests is performed, each *BUFFER_SIZE*
263  bytes in size, and with *DEPTH* requests in parallel. The first request
264  starts at the position given by *OFFSET*, each following request increases
265  the current position by *STEP_SIZE*. If *STEP_SIZE* is not given,
266  *BUFFER_SIZE* is used for its value.
267
268  If *FLUSH_INTERVAL* is specified for a write test, the request queue is
269  drained and a flush is issued before new writes are made whenever the number of
270  remaining requests is a multiple of *FLUSH_INTERVAL*. If additionally
271  ``--no-drain`` is specified, a flush is issued without draining the request
272  queue first.
273
274  if ``-i`` is specified, *AIO* option can be used to specify different
275  AIO backends: ``threads``, ``native`` or ``io_uring``.
276
277  If ``-n`` is specified, the native AIO backend is used if possible. On
278  Linux, this option only works if ``-t none`` or ``-t directsync`` is
279  specified as well.
280
281  For write tests, by default a buffer filled with zeros is written. This can be
282  overridden with a pattern byte specified by *PATTERN*.
283
284.. option:: check [--object OBJECTDEF] [--image-opts] [-q] [-f FMT] [--output=OFMT] [-r [leaks | all]] [-T SRC_CACHE] [-U] FILENAME
285
286  Perform a consistency check on the disk image *FILENAME*. The command can
287  output in the format *OFMT* which is either ``human`` or ``json``.
288  The JSON output is an object of QAPI type ``ImageCheck``.
289
290  If ``-r`` is specified, qemu-img tries to repair any inconsistencies found
291  during the check. ``-r leaks`` repairs only cluster leaks, whereas
292  ``-r all`` fixes all kinds of errors, with a higher risk of choosing the
293  wrong fix or hiding corruption that has already occurred.
294
295  Only the formats ``qcow2``, ``qed`` and ``vdi`` support
296  consistency checks.
297
298  In case the image does not have any inconsistencies, check exits with ``0``.
299  Other exit codes indicate the kind of inconsistency found or if another error
300  occurred. The following table summarizes all exit codes of the check subcommand:
301
302  0
303    Check completed, the image is (now) consistent
304  1
305    Check not completed because of internal errors
306  2
307    Check completed, image is corrupted
308  3
309    Check completed, image has leaked clusters, but is not corrupted
310  63
311    Checks are not supported by the image format
312
313  If ``-r`` is specified, exit codes representing the image state refer to the
314  state after (the attempt at) repairing it. That is, a successful ``-r all``
315  will yield the exit code 0, independently of the image state before.
316
317.. option:: commit [--object OBJECTDEF] [--image-opts] [-q] [-f FMT] [-t CACHE] [-b BASE] [-d] [-p] FILENAME
318
319  Commit the changes recorded in *FILENAME* in its base image or backing file.
320  If the backing file is smaller than the snapshot, then the backing file will be
321  resized to be the same size as the snapshot.  If the snapshot is smaller than
322  the backing file, the backing file will not be truncated.  If you want the
323  backing file to match the size of the smaller snapshot, you can safely truncate
324  it yourself once the commit operation successfully completes.
325
326  The image *FILENAME* is emptied after the operation has succeeded. If you do
327  not need *FILENAME* afterwards and intend to drop it, you may skip emptying
328  *FILENAME* by specifying the ``-d`` flag.
329
330  If the backing chain of the given image file *FILENAME* has more than one
331  layer, the backing file into which the changes will be committed may be
332  specified as *BASE* (which has to be part of *FILENAME*'s backing
333  chain). If *BASE* is not specified, the immediate backing file of the top
334  image (which is *FILENAME*) will be used. Note that after a commit operation
335  all images between *BASE* and the top image will be invalid and may return
336  garbage data when read. For this reason, ``-b`` implies ``-d`` (so that
337  the top image stays valid).
338
339.. option:: compare [--object OBJECTDEF] [--image-opts] [-f FMT] [-F FMT] [-T SRC_CACHE] [-p] [-q] [-s] [-U] FILENAME1 FILENAME2
340
341  Check if two images have the same content. You can compare images with
342  different format or settings.
343
344  The format is probed unless you specify it by ``-f`` (used for
345  *FILENAME1*) and/or ``-F`` (used for *FILENAME2*) option.
346
347  By default, images with different size are considered identical if the larger
348  image contains only unallocated and/or zeroed sectors in the area after the end
349  of the other image. In addition, if any sector is not allocated in one image
350  and contains only zero bytes in the second one, it is evaluated as equal. You
351  can use Strict mode by specifying the ``-s`` option. When compare runs in
352  Strict mode, it fails in case image size differs or a sector is allocated in
353  one image and is not allocated in the second one.
354
355  By default, compare prints out a result message. This message displays
356  information that both images are same or the position of the first different
357  byte. In addition, result message can report different image size in case
358  Strict mode is used.
359
360  Compare exits with ``0`` in case the images are equal and with ``1``
361  in case the images differ. Other exit codes mean an error occurred during
362  execution and standard error output should contain an error message.
363  The following table sumarizes all exit codes of the compare subcommand:
364
365  0
366    Images are identical
367  1
368    Images differ
369  2
370    Error on opening an image
371  3
372    Error on checking a sector allocation
373  4
374    Error on reading data
375
376.. option:: convert [--object OBJECTDEF] [--image-opts] [--target-image-opts] [--target-is-zero] [-U] [-C] [-c] [-p] [-q] [-n] [-f FMT] [-t CACHE] [-T SRC_CACHE] [-O OUTPUT_FMT] [-B BACKING_FILE] [-o OPTIONS] [-l SNAPSHOT_PARAM] [-S SPARSE_SIZE] [-m NUM_COROUTINES] [-W] FILENAME [FILENAME2 [...]] OUTPUT_FILENAME
377
378  Convert the disk image *FILENAME* or a snapshot *SNAPSHOT_PARAM*
379  to disk image *OUTPUT_FILENAME* using format *OUTPUT_FMT*. It can
380  be optionally compressed (``-c`` option) or use any format specific
381  options like encryption (``-o`` option).
382
383  Only the formats ``qcow`` and ``qcow2`` support compression. The
384  compression is read-only. It means that if a compressed sector is
385  rewritten, then it is rewritten as uncompressed data.
386
387  Image conversion is also useful to get smaller image when using a
388  growable format such as ``qcow``: the empty sectors are detected and
389  suppressed from the destination image.
390
391  *SPARSE_SIZE* indicates the consecutive number of bytes (defaults to 4k)
392  that must contain only zeros for qemu-img to create a sparse image during
393  conversion. If *SPARSE_SIZE* is 0, the source will not be scanned for
394  unallocated or zero sectors, and the destination image will always be
395  fully allocated.
396
397  You can use the *BACKING_FILE* option to force the output image to be
398  created as a copy on write image of the specified base image; the
399  *BACKING_FILE* should have the same content as the input's base image,
400  however the path, image format, etc may differ.
401
402  If a relative path name is given, the backing file is looked up relative to
403  the directory containing *OUTPUT_FILENAME*.
404
405  If the ``-n`` option is specified, the target volume creation will be
406  skipped. This is useful for formats such as ``rbd`` if the target
407  volume has already been created with site specific options that cannot
408  be supplied through qemu-img.
409
410  Out of order writes can be enabled with ``-W`` to improve performance.
411  This is only recommended for preallocated devices like host devices or other
412  raw block devices. Out of order write does not work in combination with
413  creating compressed images.
414
415  *NUM_COROUTINES* specifies how many coroutines work in parallel during
416  the convert process (defaults to 8).
417
418.. option:: create [--object OBJECTDEF] [-q] [-f FMT] [-b BACKING_FILE] [-F BACKING_FMT] [-u] [-o OPTIONS] FILENAME [SIZE]
419
420  Create the new disk image *FILENAME* of size *SIZE* and format
421  *FMT*. Depending on the file format, you can add one or more *OPTIONS*
422  that enable additional features of this format.
423
424  If the option *BACKING_FILE* is specified, then the image will record
425  only the differences from *BACKING_FILE*. No size needs to be specified in
426  this case. *BACKING_FILE* will never be modified unless you use the
427  ``commit`` monitor command (or qemu-img commit).
428
429  If a relative path name is given, the backing file is looked up relative to
430  the directory containing *FILENAME*.
431
432  Note that a given backing file will be opened to check that it is valid. Use
433  the ``-u`` option to enable unsafe backing file mode, which means that the
434  image will be created even if the associated backing file cannot be opened. A
435  matching backing file must be created or additional options be used to make the
436  backing file specification valid when you want to use an image created this
437  way.
438
439  The size can also be specified using the *SIZE* option with ``-o``,
440  it doesn't need to be specified separately in this case.
441
442
443.. option:: dd [--image-opts] [-U] [-f FMT] [-O OUTPUT_FMT] [bs=BLOCK_SIZE] [count=BLOCKS] [skip=BLOCKS] if=INPUT of=OUTPUT
444
445  dd copies from *INPUT* file to *OUTPUT* file converting it from
446  *FMT* format to *OUTPUT_FMT* format.
447
448  The data is by default read and written using blocks of 512 bytes but can be
449  modified by specifying *BLOCK_SIZE*. If count=\ *BLOCKS* is specified
450  dd will stop reading input after reading *BLOCKS* input blocks.
451
452  The size syntax is similar to :manpage:`dd(1)`'s size syntax.
453
454.. option:: info [--object OBJECTDEF] [--image-opts] [-f FMT] [--output=OFMT] [--backing-chain] [-U] FILENAME
455
456  Give information about the disk image *FILENAME*. Use it in
457  particular to know the size reserved on disk which can be different
458  from the displayed size. If VM snapshots are stored in the disk image,
459  they are displayed too.
460
461  If a disk image has a backing file chain, information about each disk image in
462  the chain can be recursively enumerated by using the option ``--backing-chain``.
463
464  For instance, if you have an image chain like:
465
466  ::
467
468    base.qcow2 <- snap1.qcow2 <- snap2.qcow2
469
470  To enumerate information about each disk image in the above chain, starting from top to base, do:
471
472  ::
473
474    qemu-img info --backing-chain snap2.qcow2
475
476  The command can output in the format *OFMT* which is either ``human`` or
477  ``json``.  The JSON output is an object of QAPI type ``ImageInfo``; with
478  ``--backing-chain``, it is an array of ``ImageInfo`` objects.
479
480  ``--output=human`` reports the following information (for every image in the
481  chain):
482
483  *image*
484    The image file name
485
486  *file format*
487    The image format
488
489  *virtual size*
490    The size of the guest disk
491
492  *disk size*
493    How much space the image file occupies on the host file system (may be
494    shown as 0 if this information is unavailable, e.g. because there is no
495    file system)
496
497  *cluster_size*
498    Cluster size of the image format, if applicable
499
500  *encrypted*
501    Whether the image is encrypted (only present if so)
502
503  *cleanly shut down*
504    This is shown as ``no`` if the image is dirty and will have to be
505    auto-repaired the next time it is opened in qemu.
506
507  *backing file*
508    The backing file name, if present
509
510  *backing file format*
511    The format of the backing file, if the image enforces it
512
513  *Snapshot list*
514    A list of all internal snapshots
515
516  *Format specific information*
517    Further information whose structure depends on the image format.  This
518    section is a textual representation of the respective
519    ``ImageInfoSpecific*`` QAPI object (e.g. ``ImageInfoSpecificQCow2``
520    for qcow2 images).
521
522.. option:: map [--object OBJECTDEF] [--image-opts] [-f FMT] [--output=OFMT] [-U] FILENAME
523
524  Dump the metadata of image *FILENAME* and its backing file chain.
525  In particular, this commands dumps the allocation state of every sector
526  of *FILENAME*, together with the topmost file that allocates it in
527  the backing file chain.
528
529  Two option formats are possible.  The default format (``human``)
530  only dumps known-nonzero areas of the file.  Known-zero parts of the
531  file are omitted altogether, and likewise for parts that are not allocated
532  throughout the chain.  ``qemu-img`` output will identify a file
533  from where the data can be read, and the offset in the file.  Each line
534  will include four fields, the first three of which are hexadecimal
535  numbers.  For example the first line of:
536
537  ::
538
539    Offset          Length          Mapped to       File
540    0               0x20000         0x50000         /tmp/overlay.qcow2
541    0x100000        0x10000         0x95380000      /tmp/backing.qcow2
542
543  means that 0x20000 (131072) bytes starting at offset 0 in the image are
544  available in /tmp/overlay.qcow2 (opened in ``raw`` format) starting
545  at offset 0x50000 (327680).  Data that is compressed, encrypted, or
546  otherwise not available in raw format will cause an error if ``human``
547  format is in use.  Note that file names can include newlines, thus it is
548  not safe to parse this output format in scripts.
549
550  The alternative format ``json`` will return an array of dictionaries
551  in JSON format.  It will include similar information in
552  the ``start``, ``length``, ``offset`` fields;
553  it will also include other more specific information:
554
555  - whether the sectors contain actual data or not (boolean field ``data``;
556    if false, the sectors are either unallocated or stored as optimized
557    all-zero clusters);
558  - whether the data is known to read as zero (boolean field ``zero``);
559  - in order to make the output shorter, the target file is expressed as
560    a ``depth``; for example, a depth of 2 refers to the backing file
561    of the backing file of *FILENAME*.
562
563  In JSON format, the ``offset`` field is optional; it is absent in
564  cases where ``human`` format would omit the entry or exit with an error.
565  If ``data`` is false and the ``offset`` field is present, the
566  corresponding sectors in the file are not yet in use, but they are
567  preallocated.
568
569  For more information, consult ``include/block/block.h`` in QEMU's
570  source code.
571
572.. option:: measure [--output=OFMT] [-O OUTPUT_FMT] [-o OPTIONS] [--size N | [--object OBJECTDEF] [--image-opts] [-f FMT] [-l SNAPSHOT_PARAM] FILENAME]
573
574  Calculate the file size required for a new image.  This information
575  can be used to size logical volumes or SAN LUNs appropriately for
576  the image that will be placed in them.  The values reported are
577  guaranteed to be large enough to fit the image.  The command can
578  output in the format *OFMT* which is either ``human`` or ``json``.
579  The JSON output is an object of QAPI type ``BlockMeasureInfo``.
580
581  If the size *N* is given then act as if creating a new empty image file
582  using ``qemu-img create``.  If *FILENAME* is given then act as if
583  converting an existing image file using ``qemu-img convert``.  The format
584  of the new file is given by *OUTPUT_FMT* while the format of an existing
585  file is given by *FMT*.
586
587  A snapshot in an existing image can be specified using *SNAPSHOT_PARAM*.
588
589  The following fields are reported:
590
591  ::
592
593    required size: 524288
594    fully allocated size: 1074069504
595
596  The ``required size`` is the file size of the new image.  It may be smaller
597  than the virtual disk size if the image format supports compact representation.
598
599  The ``fully allocated size`` is the file size of the new image once data has
600  been written to all sectors.  This is the maximum size that the image file can
601  occupy with the exception of internal snapshots, dirty bitmaps, vmstate data,
602  and other advanced image format features.
603
604.. option:: snapshot [--object OBJECTDEF] [--image-opts] [-U] [-q] [-l | -a SNAPSHOT | -c SNAPSHOT | -d SNAPSHOT] FILENAME
605
606  List, apply, create or delete snapshots in image *FILENAME*.
607
608.. option:: rebase [--object OBJECTDEF] [--image-opts] [-U] [-q] [-f FMT] [-t CACHE] [-T SRC_CACHE] [-p] [-u] -b BACKING_FILE [-F BACKING_FMT] FILENAME
609
610  Changes the backing file of an image. Only the formats ``qcow2`` and
611  ``qed`` support changing the backing file.
612
613  The backing file is changed to *BACKING_FILE* and (if the image format of
614  *FILENAME* supports this) the backing file format is changed to
615  *BACKING_FMT*. If *BACKING_FILE* is specified as "" (the empty
616  string), then the image is rebased onto no backing file (i.e. it will exist
617  independently of any backing file).
618
619  If a relative path name is given, the backing file is looked up relative to
620  the directory containing *FILENAME*.
621
622  *CACHE* specifies the cache mode to be used for *FILENAME*, whereas
623  *SRC_CACHE* specifies the cache mode for reading backing files.
624
625  There are two different modes in which ``rebase`` can operate:
626
627  Safe mode
628    This is the default mode and performs a real rebase operation. The
629    new backing file may differ from the old one and qemu-img rebase
630    will take care of keeping the guest-visible content of *FILENAME*
631    unchanged.
632
633    In order to achieve this, any clusters that differ between
634    *BACKING_FILE* and the old backing file of *FILENAME* are merged
635    into *FILENAME* before actually changing the backing file.
636
637    Note that the safe mode is an expensive operation, comparable to
638    converting an image. It only works if the old backing file still
639    exists.
640
641  Unsafe mode
642    qemu-img uses the unsafe mode if ``-u`` is specified. In this
643    mode, only the backing file name and format of *FILENAME* is changed
644    without any checks on the file contents. The user must take care of
645    specifying the correct new backing file, or the guest-visible
646    content of the image will be corrupted.
647
648    This mode is useful for renaming or moving the backing file to
649    somewhere else.  It can be used without an accessible old backing
650    file, i.e. you can use it to fix an image whose backing file has
651    already been moved/renamed.
652
653  You can use ``rebase`` to perform a "diff" operation on two
654  disk images.  This can be useful when you have copied or cloned
655  a guest, and you want to get back to a thin image on top of a
656  template or base image.
657
658  Say that ``base.img`` has been cloned as ``modified.img`` by
659  copying it, and that the ``modified.img`` guest has run so there
660  are now some changes compared to ``base.img``.  To construct a thin
661  image called ``diff.qcow2`` that contains just the differences, do:
662
663  ::
664
665    qemu-img create -f qcow2 -b modified.img diff.qcow2
666    qemu-img rebase -b base.img diff.qcow2
667
668  At this point, ``modified.img`` can be discarded, since
669  ``base.img + diff.qcow2`` contains the same information.
670
671.. option:: resize [--object OBJECTDEF] [--image-opts] [-f FMT] [--preallocation=PREALLOC] [-q] [--shrink] FILENAME [+ | -]SIZE
672
673  Change the disk image as if it had been created with *SIZE*.
674
675  Before using this command to shrink a disk image, you MUST use file system and
676  partitioning tools inside the VM to reduce allocated file systems and partition
677  sizes accordingly.  Failure to do so will result in data loss!
678
679  When shrinking images, the ``--shrink`` option must be given. This informs
680  qemu-img that the user acknowledges all loss of data beyond the truncated
681  image's end.
682
683  After using this command to grow a disk image, you must use file system and
684  partitioning tools inside the VM to actually begin using the new space on the
685  device.
686
687  When growing an image, the ``--preallocation`` option may be used to specify
688  how the additional image area should be allocated on the host.  See the format
689  description in the :ref:`notes` section which values are allowed.  Using this
690  option may result in slightly more data being allocated than necessary.
691
692.. _notes:
693
694Notes
695-----
696
697Supported image file formats:
698
699``raw``
700
701  Raw disk image format (default). This format has the advantage of
702  being simple and easily exportable to all other emulators. If your
703  file system supports *holes* (for example in ext2 or ext3 on
704  Linux or NTFS on Windows), then only the written sectors will reserve
705  space. Use ``qemu-img info`` to know the real size used by the
706  image or ``ls -ls`` on Unix/Linux.
707
708  Supported options:
709
710  ``preallocation``
711    Preallocation mode (allowed values: ``off``, ``falloc``,
712    ``full``).  ``falloc`` mode preallocates space for image by
713    calling ``posix_fallocate()``.  ``full`` mode preallocates space
714    for image by writing data to underlying storage.  This data may or
715    may not be zero, depending on the storage location.
716
717``qcow2``
718
719  QEMU image format, the most versatile format. Use it to have smaller
720  images (useful if your filesystem does not supports holes, for example
721  on Windows), optional AES encryption, zlib based compression and
722  support of multiple VM snapshots.
723
724  Supported options:
725
726  ``compat``
727    Determines the qcow2 version to use. ``compat=0.10`` uses the
728    traditional image format that can be read by any QEMU since 0.10.
729    ``compat=1.1`` enables image format extensions that only QEMU 1.1 and
730    newer understand (this is the default). Amongst others, this includes zero
731    clusters, which allow efficient copy-on-read for sparse images.
732
733  ``backing_file``
734    File name of a base image (see ``create`` subcommand)
735
736  ``backing_fmt``
737    Image format of the base image
738
739  ``encryption``
740    If this option is set to ``on``, the image is encrypted with
741    128-bit AES-CBC.
742
743    The use of encryption in qcow and qcow2 images is considered to be
744    flawed by modern cryptography standards, suffering from a number
745    of design problems:
746
747    - The AES-CBC cipher is used with predictable initialization
748      vectors based on the sector number. This makes it vulnerable to
749      chosen plaintext attacks which can reveal the existence of
750      encrypted data.
751
752    - The user passphrase is directly used as the encryption key. A
753      poorly chosen or short passphrase will compromise the security
754      of the encryption.
755
756    - In the event of the passphrase being compromised there is no way
757      to change the passphrase to protect data in any qcow images. The
758      files must be cloned, using a different encryption passphrase in
759      the new file. The original file must then be securely erased
760      using a program like shred, though even this is ineffective with
761      many modern storage technologies.
762
763    - Initialization vectors used to encrypt sectors are based on the
764      guest virtual sector number, instead of the host physical
765      sector. When a disk image has multiple internal snapshots this
766      means that data in multiple physical sectors is encrypted with
767      the same initialization vector. With the CBC mode, this opens
768      the possibility of watermarking attacks if the attack can
769      collect multiple sectors encrypted with the same IV and some
770      predictable data. Having multiple qcow2 images with the same
771      passphrase also exposes this weakness since the passphrase is
772      directly used as the key.
773
774    Use of qcow / qcow2 encryption is thus strongly discouraged. Users are
775    recommended to use an alternative encryption technology such as the
776    Linux dm-crypt / LUKS system.
777
778  ``cluster_size``
779    Changes the qcow2 cluster size (must be between 512 and
780    2M). Smaller cluster sizes can improve the image file size whereas
781    larger cluster sizes generally provide better performance.
782
783  ``preallocation``
784    Preallocation mode (allowed values: ``off``, ``metadata``,
785    ``falloc``, ``full``). An image with preallocated metadata is
786    initially larger but can improve performance when the image needs
787    to grow. ``falloc`` and ``full`` preallocations are like the same
788    options of ``raw`` format, but sets up metadata also.
789
790  ``lazy_refcounts``
791    If this option is set to ``on``, reference count updates are
792    postponed with the goal of avoiding metadata I/O and improving
793    performance. This is particularly interesting with
794    ``cache=writethrough`` which doesn't batch metadata
795    updates. The tradeoff is that after a host crash, the reference
796    count tables must be rebuilt, i.e. on the next open an (automatic)
797    ``qemu-img check -r all`` is required, which may take some time.
798
799    This option can only be enabled if ``compat=1.1`` is specified.
800
801  ``nocow``
802    If this option is set to ``on``, it will turn off COW of the file. It's
803    only valid on btrfs, no effect on other file systems.
804
805    Btrfs has low performance when hosting a VM image file, even more
806    when the guest on the VM also using btrfs as file system. Turning
807    off COW is a way to mitigate this bad performance. Generally there
808    are two ways to turn off COW on btrfs:
809
810    - Disable it by mounting with nodatacow, then all newly created files
811      will be NOCOW
812    - For an empty file, add the NOCOW file attribute. That's what this
813      option does.
814
815    Note: this option is only valid to new or empty files. If there is
816    an existing file which is COW and has data blocks already, it
817    couldn't be changed to NOCOW by setting ``nocow=on``. One can
818    issue ``lsattr filename`` to check if the NOCOW flag is set or not
819    (Capital 'C' is NOCOW flag).
820
821``Other``
822
823  QEMU also supports various other image file formats for
824  compatibility with older QEMU versions or other hypervisors,
825  including VMDK, VDI, VHD (vpc), VHDX, qcow1 and QED. For a full list
826  of supported formats see ``qemu-img --help``.  For a more detailed
827  description of these formats, see the QEMU block drivers reference
828  documentation.
829
830  The main purpose of the block drivers for these formats is image
831  conversion.  For running VMs, it is recommended to convert the disk
832  images to either raw or qcow2 in order to achieve good performance.
833