10 vzdump - Backup Utility for VMs and Containers
16 include::vzdump.1-synopsis.adoc[]
28 Backups are a requirement for any sensible IT deployment, and {pve}
29 provides a fully integrated solution, using the capabilities of each
30 storage and each guest system type. This allows the system
31 administrator to fine tune via the `mode` option between consistency
32 of the backups and downtime of the guest system.
34 {pve} backups are always full backups - containing the VM/CT
35 configuration and all data. Backups can be started via the GUI or via
36 the `vzdump` command line tool.
40 Before a backup can run, a backup storage must be defined. Refer to
41 the Storage documentation on how to add a storage. A backup storage
42 must be a file level storage, as backups are stored as regular files.
43 In most situations, using a NFS server is a good way to store backups.
44 You can save those backups later to a tape drive, for off-site
49 Backup jobs can be scheduled so that they are executed automatically
50 on specific days and times, for selectable nodes and guest systems.
51 Configuration of scheduled backups is done at the Datacenter level in
52 the GUI, which will generate a job entry in /etc/pve/jobs.cfg, which
53 will in turn be parsed and executed by the `pvescheduler` daemon.
54 These jobs use the xref:chapter_calendar_events[calendar events] for
55 defining the schedule.
57 Since scheduled backups miss their execution when the host was offline or the
58 pvescheduler was disabled during the scheduled time, it is possible to configure
59 the behaviour for catching up. By enabling the `Repeat missed` option
60 (`repeat-missed` in the config), you can tell the scheduler that it should run
61 missed jobs as soon as possible.
66 There are several ways to provide consistency (option `mode`),
67 depending on the guest type.
69 .Backup modes for VMs:
73 This mode provides the highest consistency of the backup, at the cost
74 of a short downtime in the VM operation. It works by executing an
75 orderly shutdown of the VM, and then runs a background Qemu process to
76 backup the VM data. After the backup is started, the VM goes to full
77 operation mode if it was previously running. Consistency is guaranteed
78 by using the live backup feature.
82 This mode is provided for compatibility reason, and suspends the VM
83 before calling the `snapshot` mode. Since suspending the VM results in
84 a longer downtime and does not necessarily improve the data
85 consistency, the use of the `snapshot` mode is recommended instead.
89 This mode provides the lowest operation downtime, at the cost of a
90 small inconsistency risk. It works by performing a {pve} live
91 backup, in which data blocks are copied while the VM is running. If the
92 guest agent is enabled (`agent: 1`) and running, it calls
93 `guest-fsfreeze-freeze` and `guest-fsfreeze-thaw` to improve
96 A technical overview of the {pve} live backup for QemuServer can
98 https://git.proxmox.com/?p=pve-qemu.git;a=blob_plain;f=backup.txt[here].
100 NOTE: {pve} live backup provides snapshot-like semantics on any
101 storage type. It does not require that the underlying storage supports
102 snapshots. Also please note that since the backups are done via
103 a background Qemu process, a stopped VM will appear as running for a
104 short amount of time while the VM disks are being read by Qemu.
105 However the VM itself is not booted, only its disk(s) are read.
107 .Backup modes for Containers:
111 Stop the container for the duration of the backup. This potentially
112 results in a very long downtime.
116 This mode uses rsync to copy the container data to a temporary
117 location (see option `--tmpdir`). Then the container is suspended and
118 a second rsync copies changed files. After that, the container is
119 started (resumed) again. This results in minimal downtime, but needs
120 additional space to hold the container copy.
122 When the container is on a local file system and the target storage of
123 the backup is an NFS/CIFS server, you should set `--tmpdir` to reside on a
124 local file system too, as this will result in a many fold performance
125 improvement. Use of a local `tmpdir` is also required if you want to
126 backup a local container using ACLs in suspend mode if the backup
127 storage is an NFS server.
131 This mode uses the snapshotting facilities of the underlying
132 storage. First, the container will be suspended to ensure data consistency.
133 A temporary snapshot of the container's volumes will be made and the
134 snapshot content will be archived in a tar file. Finally, the temporary
135 snapshot is deleted again.
137 NOTE: `snapshot` mode requires that all backed up volumes are on a storage that
138 supports snapshots. Using the `backup=no` mount point option individual volumes
139 can be excluded from the backup (and thus this requirement).
141 // see PVE::VZDump::LXC::prepare()
142 NOTE: By default additional mount points besides the Root Disk mount point are
143 not included in backups. For volume mount points you can set the *Backup* option
144 to include the mount point in the backup. Device and bind mounts are never
145 backed up as their content is managed outside the {pve} storage library.
150 Newer versions of vzdump encode the guest type and the
151 backup time into the filename, for example
153 vzdump-lxc-105-2009_10_09-11_04_43.tar
155 That way it is possible to store several backup in the same directory. You can
156 limit the number of backups that are kept with various retention options, see
157 the xref:vzdump_retention[Backup Retention] section below.
159 Backup File Compression
160 -----------------------
162 The backup file can be compressed with one of the following algorithms: `lzo`
163 footnote:[Lempel–Ziv–Oberhumer a lossless data compression algorithm
164 https://en.wikipedia.org/wiki/Lempel-Ziv-Oberhumer], `gzip` footnote:[gzip -
165 based on the DEFLATE algorithm https://en.wikipedia.org/wiki/Gzip] or `zstd`
166 footnote:[Zstandard a lossless data compression algorithm
167 https://en.wikipedia.org/wiki/Zstandard].
169 Currently, Zstandard (zstd) is the fastest of these three algorithms.
170 Multi-threading is another advantage of zstd over lzo and gzip. Lzo and gzip
171 are more widely used and often installed by default.
173 You can install pigz footnote:[pigz - parallel implementation of gzip
174 https://zlib.net/pigz/] as a drop-in replacement for gzip to provide better
175 performance due to multi-threading. For pigz & zstd, the amount of
176 threads/cores can be adjusted. See the
177 xref:vzdump_configuration[configuration options] below.
179 The extension of the backup file name can usually be used to determine which
180 compression algorithm has been used to create the backup.
183 |.zst | Zstandard (zstd) compression
184 |.gz or .tgz | gzip compression
185 |.lzo | lzo compression
188 If the backup file name doesn't end with one of the above file extensions, then
189 it was not compressed by vzdump.
194 For Proxmox Backup Server storages, you can optionally set up client-side
195 encryption of backups, see xref:storage_pbs_encryption[the corresponding section.]
201 With the `prune-backups` option you can specify which backups you want to keep
202 in a flexible manner. The following retention options are available:
204 `keep-all <boolean>` ::
205 Keep all backups. If this is `true`, no other options can be set.
208 Keep the last `<N>` backups.
211 Keep backups for the last `<N>` hours. If there is more than one
212 backup for a single hour, only the latest is kept.
215 Keep backups for the last `<N>` days. If there is more than one
216 backup for a single day, only the latest is kept.
219 Keep backups for the last `<N>` weeks. If there is more than one
220 backup for a single week, only the latest is kept.
222 NOTE: Weeks start on Monday and end on Sunday. The software uses the
223 `ISO week date`-system and handles weeks at the end of the year correctly.
225 `keep-monthly <N>` ::
226 Keep backups for the last `<N>` months. If there is more than one
227 backup for a single month, only the latest is kept.
230 Keep backups for the last `<N>` years. If there is more than one
231 backup for a single year, only the latest is kept.
233 The retention options are processed in the order given above. Each option
234 only covers backups within its time period. The next option does not take care
235 of already covered backups. It will only consider older backups.
237 Specify the retention options you want to use as a
238 comma-separated list, for example:
240 # vzdump 777 --prune-backups keep-last=3,keep-daily=13,keep-yearly=9
242 While you can pass `prune-backups` directly to `vzdump`, it is often more
243 sensible to configure the setting on the storage level, which can be done via
246 NOTE: The old `maxfiles` option is deprecated and should be replaced either by
247 `keep-last` or, in case `maxfiles` was `0` for unlimited retention, by
254 You can use the https://pbs.proxmox.com/docs/prune-simulator[prune simulator
255 of the Proxmox Backup Server documentation] to explore the effect of different
256 retention options with various backup schedules.
258 Retention Settings Example
259 ~~~~~~~~~~~~~~~~~~~~~~~~~~
261 The backup frequency and retention of old backups may depend on how often data
262 changes, and how important an older state may be, in a specific work load.
263 When backups act as a company's document archive, there may also be legal
264 requirements for how long backups must be kept.
266 For this example, we assume that you are doing daily backups, have a retention
267 period of 10 years, and the period between backups stored gradually grows.
269 `keep-last=3` - even if only daily backups are taken, an admin may want to
270 create an extra one just before or after a big upgrade. Setting keep-last
273 `keep-hourly` is not set - for daily backups this is not relevant. You cover
274 extra manual backups already, with keep-last.
276 `keep-daily=13` - together with keep-last, which covers at least one
277 day, this ensures that you have at least two weeks of backups.
279 `keep-weekly=8` - ensures that you have at least two full months of
282 `keep-monthly=11` - together with the previous keep settings, this
283 ensures that you have at least a year of monthly backups.
285 `keep-yearly=9` - this is for the long term archive. As you covered the
286 current year with the previous options, you would set this to nine for the
287 remaining ones, giving you a total of at least 10 years of coverage.
289 We recommend that you use a higher retention period than is minimally required
290 by your environment; you can always reduce it if you find it is unnecessarily
291 high, but you cannot recreate backups once they have been removed.
293 [[vzdump_protection]]
297 You can mark a backup as `protected` to prevent its removal. Attempting to
298 remove a protected backup via {pve}'s UI, CLI or API will fail. However, this
299 is enforced by {pve} and not the file-system, that means that a manual removal
300 of a backup file itself is still possible for anyone with write access to the
301 underlying backup storage.
303 NOTE: Protected backups are ignored by pruning and do not count towards the
306 For filesystem-based storages, the protection is implemented via a sentinel file
307 `<backup-name>.protected`. For Proxmox Backup Server, it is handled on the
308 server side (available since Proxmox Backup Server version 2.1).
310 Use the storage option `max-protected-backups` to control how many protected
311 backups per guest are allowed on the storage. Use `-1` for unlimited. The
312 default is unlimited for users with `Datastore.Allocate` privilege and `5` for
319 You can add notes to backups using the 'Edit Notes' button in the UI or via the
322 It is also possible to specify a template for generating notes dynamically for
323 a backup job and for manual backup. The template string can contain variables,
324 surrounded by two curly braces, which will be replaced by the corresponding
325 value when the backup is executed.
327 Currently supported are:
329 * `{{cluster}}` the cluster name, if any
330 * `{{guestname}}` the virtual guest's assigned name
331 * `{{node}}` the host name of the node the backup is being created
332 * `{{vmid}}` the numerical VMID of the guest
334 When specified via API or CLI, it needs to be a single line, where newline and
335 backslash need to be escaped as literal `\n` and `\\` respectively.
341 A backup archive can be restored through the {pve} web GUI or through the
345 `pct restore`:: Container restore utility
347 `qmrestore`:: Virtual Machine restore utility
349 For details see the corresponding manual pages.
354 Restoring one or more big backups may need a lot of resources, especially
355 storage bandwidth for both reading from the backup storage and writing to
356 the target storage. This can negatively affect other virtual guests as access
357 to storage can get congested.
359 To avoid this you can set bandwidth limits for a backup job. {pve}
360 implements two kinds of limits for restoring and archive:
362 * per-restore limit: denotes the maximal amount of bandwidth for
363 reading from a backup archive
365 * per-storage write limit: denotes the maximal amount of bandwidth used for
366 writing to a specific storage
368 The read limit indirectly affects the write limit, as we cannot write more
369 than we read. A smaller per-job limit will overwrite a bigger per-storage
370 limit. A bigger per-job limit will only overwrite the per-storage limit if
371 you have `Data.Allocate' permissions on the affected storage.
373 You can use the `--bwlimit <integer>` option from the restore CLI commands
374 to set up a restore job specific bandwidth limit. Kibit/s is used as unit
375 for the limit, this means passing `10240' will limit the read speed of the
376 backup to 10 MiB/s, ensuring that the rest of the possible storage bandwidth
377 is available for the already running virtual guests, and thus the backup
378 does not impact their operations.
380 NOTE: You can use `0` for the `bwlimit` parameter to disable all limits for
381 a specific restore job. This can be helpful if you need to restore a very
382 important virtual guest as fast as possible. (Needs `Data.Allocate'
383 permissions on storage)
385 Most times your storage's generally available bandwidth stays the same over
386 time, thus we implemented the possibility to set a default bandwidth limit
387 per configured storage, this can be done with:
390 # pvesm set STORAGEID --bwlimit restore=KIBs
396 Restoring a large backup can take a long time, in which a guest is still
397 unavailable. For VM backups stored on a Proxmox Backup Server, this wait
398 time can be mitigated using the live-restore option.
400 Enabling live-restore via either the checkbox in the GUI or the `--live-restore`
401 argument of `qmrestore` causes the VM to start as soon as the restore
402 begins. Data is copied in the background, prioritizing chunks that the VM is
405 Note that this comes with two caveats:
407 * During live-restore, the VM will operate with limited disk read speeds, as
408 data has to be loaded from the backup server (once loaded, it is immediately
409 available on the destination storage however, so accessing data twice only
410 incurs the penalty the first time). Write speeds are largely unaffected.
411 * If the live-restore fails for any reason, the VM will be left in an
412 undefined state - that is, not all data might have been copied from the
413 backup, and it is _most likely_ not possible to keep any data that was written
414 during the failed restore operation.
416 This mode of operation is especially useful for large VMs, where only a small
417 amount of data is required for initial operation, e.g. web servers - once the OS
418 and necessary services have been started, the VM is operational, while the
419 background task continues copying seldom used data.
424 The 'File Restore' button in the 'Backups' tab of the storage GUI can be used to
425 open a file browser directly on the data contained in a backup. This feature
426 is only available for backups on a Proxmox Backup Server.
428 For containers, the first layer of the file tree shows all included 'pxar'
429 archives, which can be opened and browsed freely. For VMs, the first layer shows
430 contained drive images, which can be opened to reveal a list of supported
431 storage technologies found on the drive. In the most basic case, this will be an
432 entry called 'part', representing a partition table, which contains entries for
433 each partition found on the drive. Note that for VMs, not all data might be
434 accessible (unsupported guest file systems, storage technologies, etc...).
436 Files and directories can be downloaded using the 'Download' button, the latter
437 being compressed into a zip archive on the fly.
439 To enable secure access to VM images, which might contain untrusted data, a
440 temporary VM (not visible as a guest) is started. This does not mean that data
441 downloaded from such an archive is inherently safe, but it avoids exposing the
442 hypervisor system to danger. The VM will stop itself after a timeout. This
443 entire process happens transparently from a user's point of view.
445 [[vzdump_configuration]]
449 Global configuration is stored in `/etc/vzdump.conf`. The file uses a
450 simple colon separated key/value format. Each line has the following
455 Blank lines in the file are ignored, and lines starting with a `#`
456 character are treated as comments and are also ignored. Values from
457 this file are used as default, and can be overwritten on the command
460 We currently support the following options:
462 include::vzdump.conf.5-opts.adoc[]
465 .Example `vzdump.conf` Configuration
467 tmpdir: /mnt/fast_local_disk
468 storage: my_backup_storage
476 You can specify a hook script with option `--script`. This script is
477 called at various phases of the backup process, with parameters
478 accordingly set. You can find an example in the documentation
479 directory (`vzdump-hook-script.pl`).
484 NOTE: this option is only available for container backups.
486 `vzdump` skips the following files by default (disable with the option
493 You can also manually specify (additional) exclude paths, for example:
495 # vzdump 777 --exclude-path /tmp/ --exclude-path '/var/foo*'
497 excludes the directory `/tmp/` and any file or directory named `/var/foo`,
498 `/var/foobar`, and so on.
500 Paths that do not start with a `/` are not anchored to the container's root,
501 but will match relative to any subdirectory. For example:
503 # vzdump 777 --exclude-path bar
505 excludes any file or directory named `/bar`, `/var/bar`, `/var/foo/bar`, and
506 so on, but not `/bar2`.
508 Configuration files are also stored inside the backup archive
509 (in `./etc/vzdump/`) and will be correctly restored.
514 Simply dump guest 777 - no snapshot, just archive the guest private area and
515 configuration files to the default dump directory (usually
516 `/var/lib/vz/dump/`).
520 Use rsync and suspend/resume to create a snapshot (minimal downtime).
522 # vzdump 777 --mode suspend
524 Backup all guest systems and send notification mails to root and admin.
526 # vzdump --all --mode suspend --mailto root --mailto admin
528 Use snapshot mode (no downtime) and non-default dump directory.
530 # vzdump 777 --dumpdir /mnt/backup --mode snapshot
532 Backup more than one guest (selectively)
534 # vzdump 101 102 103 --mailto root
536 Backup all guests excluding 101 and 102
538 # vzdump --mode suspend --exclude 101,102
540 Restore a container to a new CT 600
542 # pct restore 600 /mnt/backup/vzdump-lxc-777.tar
544 Restore a QemuServer VM to VM 601
546 # qmrestore /mnt/backup/vzdump-qemu-888.vma 601
548 Clone an existing container 101 to a new container 300 with a 4GB root
549 file system, using pipes
551 # vzdump 101 --stdout | pct restore --rootfs 4 300 -
555 include::pve-copyright.adoc[]