1BORG(1) Borg - Deduplicating Archiver BORG(1)
2
6 borg - BorgBackup is a deduplicating backup program with optional com‐
7 pression and authenticated encryption.
8
10 Borg consists of a number of commands. Each command accepts a number of
11 arguments and options and interprets various environment variables.
12 The following sections will describe each command in detail.
13 Commands, options, parameters, paths and such are set in fixed-width.
15 Option values are underlined. Borg has few options accepting a fixed
16 set of values (e.g. --encryption of borg_init). Experimental features
17 are marked with red stripes on the sides, like this paragraph.
18 Experimental features are not stable, which means that they may be
20 changed in incompatible ways or even removed entirely without prior
21 notice in following releases.
22 Positional Arguments and Options: Order matters
24 Borg only supports taking options (-s and --progress in the example) to
25 the left or right of all positional arguments (repo::archive and path
26 in the example), but not in between them:
27 borg create -s --progress repo::archive path # good and preferred
29 borg create repo::archive path -s --progress # also works
30 borg create -s repo::archive path --progress # works, but ugly
31 borg create repo::archive -s --progress path # BAD
32 This is due to a problem in the argparse module:
34 http://bugs.python.org/issue15112
35 Repository URLs
37 Local filesystem (or locally mounted network filesystem):
38 /path/to/repo - filesystem path to repo directory, absolute path
40 path/to/repo - filesystem path to repo directory, relative path
42 Also, stuff like ~/path/to/repo or ~other/path/to/repo works (this is
44 expanded by your shell).
45 Note: you may also prepend a file:// to a filesystem path to get URL
47 style.
48 Remote repositories accessed via ssh user@host:
50 user@host:/path/to/repo - remote repo, absolute path
52 ssh://user@host:port/path/to/repo - same, alternative syntax, port can
54 be given
55 Remote repositories with relative paths can be given using this syntax:
57 user@host:path/to/repo - path relative to current directory
59 user@host:~/path/to/repo - path relative to user's home directory
61 user@host:~other/path/to/repo - path relative to other's home directory
63 Note: giving user@host:/./path/to/repo or user@host:/~/path/to/repo or
65 user@host:/~other/path/to/repo is also supported, but not required
66 here.
67 Remote repositories with relative paths, alternative syntax with port:
69 ssh://user@host:port/./path/to/repo - path relative to current direc‐
71 tory
72 ssh://user@host:port/~/path/to/repo - path relative to user's home
74 directory
75 ssh://user@host:port/~other/path/to/repo - path relative to other's
77 home directory
78 If you frequently need the same repo URL, it is a good idea to set the
80 BORG_REPO environment variable to set a default for the repo URL:
81 export BORG_REPO='ssh://user@host:port/path/to/repo'
83 Then just leave away the repo URL if only a repo URL is needed and you
85 want to use the default - it will be read from BORG_REPO then.
86 Use :: syntax to give the repo URL when syntax requires giving a posi‐
88 tional argument for the repo (e.g. borg mount :: /mnt).
89 Repository / Archive Locations
91 Many commands want either a repository (just give the repo URL, see
92 above) or an archive location, which is a repo URL followed by ::ar‐
93 chive_name.
94 Archive names must not contain the / (slash) character. For simplicity,
96 maybe also avoid blanks or other characters that have special meaning
97 on the shell or in a filesystem (borg mount will use the archive name
98 as directory name).
99 If you have set BORG_REPO (see above) and an archive location is
101 needed, use ::archive_name - the repo URL part is then read from
102 BORG_REPO.
103 Logging
105 Borg writes all log output to stderr by default. But please note that
106 something showing up on stderr does not indicate an error condition
107 just because it is on stderr. Please check the log levels of the mes‐
108 sages and the return code of borg for determining error, warning or
109 success conditions.
110 If you want to capture the log output to a file, just redirect it:
112 borg create repo::archive myfiles 2>> logfile
114 Custom logging configurations can be implemented via BORG_LOGGING_CONF.
116 The log level of the builtin logging configuration defaults to WARNING.
118 This is because we want Borg to be mostly silent and only output warn‐
119 ings, errors and critical messages, unless output has been requested by
120 supplying an option that implies output (e.g. --list or --progress).
121 Log levels: DEBUG < INFO < WARNING < ERROR < CRITICAL
123 Use --debug to set DEBUG log level - to get debug, info, warning, error
125 and critical level output.
126 Use --info (or -v or --verbose) to set INFO log level - to get info,
128 warning, error and critical level output.
129 Use --warning (default) to set WARNING log level - to get warning,
131 error and critical level output.
132 Use --error to set ERROR log level - to get error and critical level
134 output.
135 Use --critical to set CRITICAL log level - to get critical level out‐
137 put.
138 While you can set misc. log levels, do not expect that every command
140 will give different output on different log levels - it's just a possi‐
141 bility.
142 WARNING:
144 Options --critical and --error are provided for completeness, their
145 usage is not recommended as you might miss important information.
146 Return codes
148 Borg can exit with the following return codes (rc):
149 ┌────────────┬────────────────────────────┐
151 │Return code │ Meaning │
152 ├────────────┼────────────────────────────┤
153 │0 │ success (logged as INFO) │
154 ├────────────┼────────────────────────────┤
155 │1 │ warning (operation reached │
156 │ │ its normal end, but there │
157 │ │ were warnings -- you │
158 │ │ should check the log, │
159 │ │ logged as WARNING) │
160 ├────────────┼────────────────────────────┤
161 │2 │ error (like a fatal error, │
162 │ │ a local or remote excep‐ │
163 │ │ tion, the operation did │
164 │ │ not reach its normal end, │
165 │ │ logged as ERROR) │
166 ├────────────┼────────────────────────────┤
167 │128+N │ killed by signal N (e.g. │
168 │ │ 137 == kill -9) │
169 └────────────┴────────────────────────────┘
170 If you use --show-rc, the return code is also logged at the indicated
172 level as the last log entry.
173 Environment Variables
175 Borg uses some environment variables for automation:
176 General:
178 BORG_REPO
180 When set, use the value to give the default repository
181 location. If a command needs an archive parameter, you
182 can abbreviate as ::archive. If a command needs a reposi‐
183 tory parameter, you can either leave it away or abbrevi‐
184 ate as ::, if a positional parameter is required.
185 BORG_PASSPHRASE
187 When set, use the value to answer the passphrase question
188 for encrypted repositories. It is used when a passphrase
189 is needed to access an encrypted repo as well as when a
190 new passphrase should be initially set when initializing
191 an encrypted repo. See also BORG_NEW_PASSPHRASE.
192 BORG_PASSCOMMAND
194 When set, use the standard output of the command (trail‐
195 ing newlines are stripped) to answer the passphrase ques‐
196 tion for encrypted repositories. It is used when a
197 passphrase is needed to access an encrypted repo as well
198 as when a new passphrase should be initially set when
199 initializing an encrypted repo. Note that the command is
200 executed without a shell. So variables, like $HOME will
201 work, but ~ won't. If BORG_PASSPHRASE is also set, it
202 takes precedence. See also BORG_NEW_PASSPHRASE.
203 BORG_PASSPHRASE_FD
205 When set, specifies a file descriptor to read a
206 passphrase from. Programs starting borg may choose to
207 open an anonymous pipe and use it to pass a passphrase.
208 This is safer than passing via BORG_PASSPHRASE, because
209 on some systems (e.g. Linux) environment can be examined
210 by other processes. If BORG_PASSPHRASE or BORG_PASSCOM‐
211 MAND are also set, they take precedence.
212 BORG_NEW_PASSPHRASE
214 When set, use the value to answer the passphrase question
215 when a new passphrase is asked for. This variable is
216 checked first. If it is not set, BORG_PASSPHRASE and
217 BORG_PASSCOMMAND will also be checked. Main usecase for
218 this is to fully automate borg change-passphrase.
219 BORG_DISPLAY_PASSPHRASE
221 When set, use the value to answer the "display the
222 passphrase for verification" question when defining a new
223 passphrase for encrypted repositories.
224 BORG_HOSTNAME_IS_UNIQUE=no
226 Borg assumes that it can derive a unique hostname / iden‐
227 tity (see borg debug info). If this is not the case or
228 you do not want Borg to automatically remove stale locks,
229 set this to no.
230 BORG_HOST_ID
232 Borg usually computes a host id from the FQDN plus the
233 results of uuid.getnode() (which usually returns a unique
234 id based on the MAC address of the network interface.
235 Except if that MAC happens to be all-zero - in that case
236 it returns a random value, which is not what we want
237 (because it kills automatic stale lock removal). So, if
238 you have a all-zero MAC address or other reasons to bet‐
239 ter externally control the host id, just set this envi‐
240 ronment variable to a unique value. If all your FQDNs are
241 unique, you can just use the FQDN. If not, use
242 fqdn@uniqueid.
243 BORG_LOGGING_CONF
245 When set, use the given filename as INI-style logging
246 configuration. A basic example conf can be found at
247 docs/misc/logging.conf.
248 BORG_RSH
250 When set, use this command instead of ssh. This can be
251 used to specify ssh options, such as a custom identity
252 file ssh -i /path/to/private/key. See man ssh for other
253 options. Using the --rsh CMD commandline option overrides
254 the environment variable.
255 BORG_REMOTE_PATH
257 When set, use the given path as borg executable on the
258 remote (defaults to "borg" if unset). Using
259 --remote-path PATH commandline option overrides the envi‐
260 ronment variable.
261 BORG_FILES_CACHE_TTL
263 When set to a numeric value, this determines the maximum
264 "time to live" for the files cache entries (default: 20).
265 The files cache is used to quickly determine whether a
266 file is unchanged. The FAQ explains this more detailed
267 in: always_chunking
268 BORG_SHOW_SYSINFO
270 When set to no (default: yes), system information (like
271 OS, Python version, ...) in exceptions is not shown.
272 Please only use for good reasons as it makes issues
273 harder to analyze.
274 BORG_WORKAROUNDS
276 A list of comma separated strings that trigger work‐
277 arounds in borg, e.g. to work around bugs in other soft‐
278 ware.
279 Currently known strings are:
281 basesyncfile
283 Use the more simple BaseSyncFile code to avoid
284 issues with sync_file_range. You might need this
285 to run borg on WSL (Windows Subsystem for Linux)
286 or in systemd.nspawn containers on some architec‐
287 tures (e.g. ARM). Using this does not affect data
288 safety, but might result in a more bursty write to
289 disk behaviour (not continuously streaming to
290 disk).
291 TMPDIR where temporary files are stored (might need a lot of
293 temporary space for some operations), see tempfile for
294 details
295 Some automatic answerers (if set, they automatically answer confirma‐
297 tion questions):
298 BORG_UNKNOWN_UNENCRYPTED_REPO_ACCESS_IS_OK=no (or =yes)
300 For "Warning: Attempting to access a previously unknown
301 unencrypted repository"
302 BORG_RELOCATED_REPO_ACCESS_IS_OK=no (or =yes)
304 For "Warning: The repository at location ... was previ‐
305 ously located at ..."
306 BORG_CHECK_I_KNOW_WHAT_I_AM_DOING=NO (or =YES)
308 For "Warning: 'check --repair' is an experimental feature
309 that might result in data loss."
310 BORG_DELETE_I_KNOW_WHAT_I_AM_DOING=NO (or =YES)
312 For "You requested to completely DELETE the repository
313 including all archives it contains:"
314 BORG_RECREATE_I_KNOW_WHAT_I_AM_DOING=NO (or =YES)
316 For "recreate is an experimental feature."
317 Note: answers are case sensitive. setting an invalid answer
319 value might either give the default answer or ask you interac‐
320 tively, depending on whether retries are allowed (they by
321 default are allowed). So please test your scripts interactively
322 before making them a non-interactive script.
323 Directories and files:
325 BORG_BASE_DIR
327 Default to '$HOME', '~$USER', '~' (in that order)'. If
328 we refer to ~ below, we in fact mean BORG_BASE_DIR.
329 BORG_CACHE_DIR
331 Default to '~/.cache/borg'. This directory contains the
332 local cache and might need a lot of space for dealing
333 with big repositories. Make sure you're aware of the
334 associated security aspects of the cache location:
335 cache_security
336 BORG_CONFIG_DIR
338 Default to '~/.config/borg'. This directory contains the
339 whole config directories.
340 BORG_SECURITY_DIR
342 Default to '~/.config/borg/security'. This directory con‐
343 tains information borg uses to track its usage of NONCES
344 ("numbers used once" - usually in encryption context) and
345 other security relevant data. Will move with BORG_CON‐
346 FIG_DIR variable unless specified.
347 BORG_KEYS_DIR
349 Default to '~/.config/borg/keys'. This directory contains
350 keys for encrypted repositories.
351 BORG_KEY_FILE
353 When set, use the given filename as repository key file.
354 Building:
356 BORG_OPENSSL_PREFIX
358 Adds given OpenSSL header file directory to the default
359 locations (setup.py).
360 BORG_LIBLZ4_PREFIX
362 Adds given prefix directory to the default locations. If
363 a 'include/lz4.h' is found Borg will be linked against
364 the system liblz4 instead of a bundled implementation.
365 (setup.py)
366 BORG_LIBB2_PREFIX
368 Adds given prefix directory to the default locations. If
369 a 'include/blake2.h' is found Borg will be linked against
370 the system libb2 instead of a bundled implementation.
371 (setup.py)
372 BORG_LIBZSTD_PREFIX
374 Adds given prefix directory to the default locations. If
375 a 'include/zstd.h' is found Borg will be linked against
376 the system libzstd instead of a bundled implementation.
377 (setup.py)
378 Please note:
380 · be very careful when using the "yes" sayers, the warnings with prompt
382 exist for your / your data's security/safety
383 · also be very careful when putting your passphrase into a script, make
385 sure it has appropriate file permissions (e.g. mode 600, root:root).
386 File systems
388 We strongly recommend against using Borg (or any other database-like
389 software) on non-journaling file systems like FAT, since it is not pos‐
390 sible to assume any consistency in case of power failures (or a sudden
391 disconnect of an external drive or similar failures).
392 While Borg uses a data store that is resilient against these failures
394 when used on journaling file systems, it is not possible to guarantee
395 this with some hardware -- independent of the software used. We don't
396 know a list of affected hardware.
397 If you are suspicious whether your Borg repository is still consistent
399 and readable after one of the failures mentioned above occurred, run
400 borg check --verify-data to make sure it is consistent.
401 Requirements for Borg repository file systems
403 · Long file names
405 · At least three directory levels with short names
407 · Typically, file sizes up to a few hundred MB. Large repositories may
409 require large files (>2 GB).
410 · Up to 1000 files per directory (10000 for repositories initialized
412 with Borg 1.0)
413 · mkdir(2) should be atomic, since it is used for locking
415 · Hardlinks are needed for borg_upgrade --inplace
417 Units
419 To display quantities, Borg takes care of respecting the usual conven‐
420 tions of scale. Disk sizes are displayed in decimal, using powers of
421 ten (so kB means 1000 bytes). For memory usage, binary prefixes are
422 used, and are indicated using the IEC binary prefixes, using powers of
423 two (so KiB means 1024 bytes).
424 Date and Time
426 We format date and time conforming to ISO-8601, that is: YYYY-MM-DD and
427 HH:MM:SS (24h clock).
428 For more information about that, see: https://xkcd.com/1179/
430 Unless otherwise noted, we display local date and time. Internally, we
432 store and process date and time as UTC.
433 Resource Usage
435 Borg might use a lot of resources depending on the size of the data set
436 it is dealing with.
437 If one uses Borg in a client/server way (with a ssh: repository), the
439 resource usage occurs in part on the client and in another part on the
440 server.
441 If one uses Borg as a single process (with a filesystem repo), all the
443 resource usage occurs in that one process, so just add up client +
444 server to get the approximate resource usage.
445 CPU client:
447 · borg create: does chunking, hashing, compression, crypto (high
449 CPU usage)
450 · chunks cache sync: quite heavy on CPU, doing lots of hashtable
452 operations.
453 · borg extract: crypto, decompression (medium to high CPU usage)
455 · borg check: similar to extract, but depends on options given.
457 · borg prune / borg delete archive: low to medium CPU usage
459 · borg delete repo: done on the server
461 It won't go beyond 100% of 1 core as the code is currently sin‐
463 gle-threaded. Especially higher zlib and lzma compression lev‐
464 els use significant amounts of CPU cycles. Crypto might be cheap
465 on the CPU (if hardware accelerated) or expensive (if not).
466 CPU server:
468 It usually doesn't need much CPU, it just deals with the
469 key/value store (repository) and uses the repository index for
470 that.
471 borg check: the repository check computes the checksums of all
473 chunks (medium CPU usage) borg delete repo: low CPU usage
474 CPU (only for client/server operation):
476 When using borg in a client/server way with a ssh:-type repo,
477 the ssh processes used for the transport layer will need some
478 CPU on the client and on the server due to the crypto they are
479 doing - esp. if you are pumping big amounts of data.
480 Memory (RAM) client:
482 The chunks index and the files index are read into memory for
483 performance reasons. Might need big amounts of memory (see
484 below). Compression, esp. lzma compression with high levels
485 might need substantial amounts of memory.
486 Memory (RAM) server:
488 The server process will load the repository index into memory.
489 Might need considerable amounts of memory, but less than on the
490 client (see below).
491 Chunks index (client only):
493 Proportional to the amount of data chunks in your repo. Lots of
494 chunks in your repo imply a big chunks index. It is possible to
495 tweak the chunker params (see create options).
496 Files index (client only):
498 Proportional to the amount of files in your last backups. Can be
499 switched off (see create options), but next backup might be much
500 slower if you do. The speed benefit of using the files cache is
501 proportional to file size.
502 Repository index (server only):
504 Proportional to the amount of data chunks in your repo. Lots of
505 chunks in your repo imply a big repository index. It is possi‐
506 ble to tweak the chunker params (see create options) to influ‐
507 ence the amount of chunks being created.
508 Temporary files (client):
510 Reading data and metadata from a FUSE mounted repository will
511 consume up to the size of all deduplicated, small chunks in the
512 repository. Big chunks won't be locally cached.
513 Temporary files (server):
515 A non-trivial amount of data will be stored on the remote temp
516 directory for each client that connects to it. For some remotes,
517 this can fill the default temporary directory at /tmp. This can
518 be remediated by ensuring the $TMPDIR, $TEMP, or $TMP environ‐
519 ment variable is properly set for the sshd process. For some
520 OSes, this can be done just by setting the correct value in the
521 other cases it may be neccessary to first enable PermitUserEnvi‐
522 ronment yes in your sshd_config file, then add environ‐
523 ment="TMPDIR=/my/big/tmpdir" at the start of the public key to
524 be used in the authorized_hosts file.
525 Cache files (client only):
527 Contains the chunks index and files index (plus a collection of
528 single- archive chunk indexes which might need huge amounts of
529 disk space, depending on archive count and size - see FAQ about
530 how to reduce).
531 Network (only for client/server operation):
533 If your repository is remote, all deduplicated (and optionally
534 compressed/ encrypted) data of course has to go over the connec‐
535 tion (ssh:// repo url). If you use a locally mounted network
536 filesystem, additionally some copy operations used for transac‐
537 tion support also go over the connection. If you backup multiple
538 sources to one target repository, additional traffic happens for
539 cache resynchronization.
540 Support for file metadata
542 Besides regular file and directory structures, Borg can preserve
543 · symlinks (stored as symlink, the symlink is not followed)
545 · special files:
547 · character and block device files (restored via mknod)
549 · FIFOs ("named pipes")
551 · special file contents can be backed up in --read-special mode. By
553 default the metadata to create them with mknod(2), mkfifo(2) etc.
554 is stored.
555 · hardlinked regular files, devices, FIFOs (considering all items in
557 the same archive)
558 · timestamps in nanosecond precision: mtime, atime, ctime
560 · other timestamps: birthtime (on platforms supporting it)
562 · permissions:
564 · IDs of owning user and owning group
566 · names of owning user and owning group (if the IDs can be resolved)
568 · Unix Mode/Permissions (u/g/o permissions, suid, sgid, sticky)
570 On some platforms additional features are supported:
572 ┌─────────────────┬──────────┬───────────┬───────────┐
574 │Platform │ ACLs [5] │ xattr [6] │ Flags [7] │
575 ├─────────────────┼──────────┼───────────┼───────────┤
576 │Linux │ Yes │ Yes │ Yes [1] │
577 ├─────────────────┼──────────┼───────────┼───────────┤
578 │Mac OS X │ Yes │ Yes │ Yes (all) │
579 ├─────────────────┼──────────┼───────────┼───────────┤
580 │FreeBSD │ Yes │ Yes │ Yes (all) │
581 ├─────────────────┼──────────┼───────────┼───────────┤
582 │OpenBSD │ n/a │ n/a │ Yes (all) │
583 ├─────────────────┼──────────┼───────────┼───────────┤
584 │NetBSD │ n/a │ No [2] │ Yes (all) │
585 ├─────────────────┼──────────┼───────────┼───────────┤
586 │Solaris and de‐ │ No [3] │ No [3] │ n/a │
587 │rivatives │ │ │ │
588 ├─────────────────┼──────────┼───────────┼───────────┤
589 │Windows (cygwin) │ No [4] │ No │ No │
590 └─────────────────┴──────────┴───────────┴───────────┘
591 Other Unix-like operating systems may work as well, but have not been
593 tested at all.
594 Note that most of the platform-dependent features also depend on the
596 file system. For example, ntfs-3g on Linux isn't able to convey NTFS
597 ACLs.
598 [1] Only "nodump", "immutable", "compressed" and "append" are sup‐
600 ported. Feature request #618 for more flags.
601 [2] Feature request #1332
603 [3] Feature request #1337
605 [4] Cygwin tries to map NTFS ACLs to permissions with varying degrees
607 of success.
608 [5] The native access control list mechanism of the OS. This normally
610 limits access to non-native ACLs. For example, NTFS ACLs aren't
611 completely accessible on Linux with ntfs-3g.
612 [6] extended attributes; key-value pairs attached to a file, mainly
614 used by the OS. This includes resource forks on Mac OS X.
615 [7] aka BSD flags. The Linux set of flags [1] is portable across plat‐
617 forms. The BSDs define additional flags.
618 In case you are interested in more details (like formulas), please
620 see internals. For details on the available JSON output, refer to
621 json_output.
622 Common options
624 All Borg commands share these options:
625 -h, --help
627 show this help message and exit
628 --critical
630 work on log level CRITICAL
631 --error
633 work on log level ERROR
634 --warning
636 work on log level WARNING (default)
637 --info, -v, --verbose
639 work on log level INFO
640 --debug
642 enable debug output, work on log level DEBUG
643 --debug-topic TOPIC
645 enable TOPIC debugging (can be specified multiple times). The
646 logger path is borg.debug.<TOPIC> if TOPIC is not fully quali‐
647 fied.
648 -p, --progress
650 show progress information
651 --log-json
653 Output one JSON object per log line instead of formatted text.
654 --lock-wait SECONDS
656 wait at most SECONDS for acquiring a repository/cache lock
657 (default: 1).
658 --show-version
660 show/log the borg version
661 --show-rc
663 show/log the return code (rc)
664 --umask M
666 set umask to M (local and remote, default: 0077)
667 --remote-path PATH
669 use PATH as borg executable on the remote (default: "borg")
670 --remote-ratelimit RATE
672 set remote network upload rate limit in kiByte/s (default:
673 0=unlimited)
674 --consider-part-files
676 treat part files like normal files (e.g. to list/extract them)
677 --debug-profile FILE
679 Write execution profile in Borg format into FILE. For local use
680 a Python-compatible file can be generated by suffixing FILE with
681 ".pyprof".
682 --rsh RSH
684 Use this command to connect to the 'borg serve' process
685 (default: 'ssh')
686 Examples
688 # Create an archive and log: borg version, files list, return code
689 $ borg create --show-version --list --show-rc /path/to/repo::my-files files
690
692 borg [common options] init [options] [REPOSITORY]
693 Description
695 This command initializes an empty repository. A repository is a
696 filesystem directory containing the deduplicated data from zero or more
697 archives.
698 Encryption can be enabled at repository init time. It cannot be changed
700 later.
701 It is not recommended to work without encryption. Repository encryption
703 protects you e.g. against the case that an attacker has access to your
704 backup repository.
705 But be careful with the key / the passphrase:
707 If you want "passphrase-only" security, use one of the repokey modes.
709 The key will be stored inside the repository (in its "config" file). In
710 above mentioned attack scenario, the attacker will have the key (but
711 not the passphrase).
712 If you want "passphrase and having-the-key" security, use one of the
714 keyfile modes. The key will be stored in your home directory (in .con‐
715 fig/borg/keys). In the attack scenario, the attacker who has just
716 access to your repo won't have the key (and also not the passphrase).
717 Make a backup copy of the key file (keyfile mode) or repo config file
719 (repokey mode) and keep it at a safe place, so you still have the key
720 in case it gets corrupted or lost. Also keep the passphrase at a safe
721 place. The backup that is encrypted with that key won't help you with
722 that, of course.
723 Make sure you use a good passphrase. Not too short, not too simple. The
725 real encryption / decryption key is encrypted with / locked by your
726 passphrase. If an attacker gets your key, he can't unlock and use it
727 without knowing the passphrase.
728 Be careful with special or non-ascii characters in your passphrase:
730 · Borg processes the passphrase as unicode (and encodes it as utf-8),
732 so it does not have problems dealing with even the strangest charac‐
733 ters.
734 · BUT: that does not necessarily apply to your OS / VM / keyboard con‐
736 figuration.
737 So better use a long passphrase made from simple ascii chars than one
739 that includes non-ascii stuff or characters that are hard/impossible to
740 enter on a different keyboard layout.
741 You can change your passphrase for existing repos at any time, it won't
743 affect the encryption/decryption key or other secrets.
744 Encryption modes
746 ────────────────────────────────────────────────────────────────────
747 Hash/MAC Not encrypted no Not encrypted, Encrypted (AEAD
748 auth but authenti‐ w/ AES) and
749 cated authenticated
750 ────────────────────────────────────────────────────────────────────
751 SHA-256 none authenticated repokey keyfile
752 ────────────────────────────────────────────────────────────────────
753 BLAKE2b n/a authenti‐ repokey-blake2
754 cated-blake2 keyfile-blake2
755 ┌─────────┬──────────────────┬──────────────────┬──────────────────┐
756 │ │ │ │ │
757 Ma│rked mode│s are new in Borg 1│.1 and are not back│wards-compatible wi│th
758 Bo│rg 1.0.x. │ │ │ │
759 │ │ │ │ │
760 On│modern In│tel/AMD CPUs (excep│t very cheap ones),│AES is usually har│d‐
761 wa│re-acceler│ated. BLAKE2b is│ faster than SHA25│6 on Intel/AMD 64-b│it
762 CP│Us (except│AMD Ryzen and futu│re CPUs with SHA│ extensions), whi│ch
763 ma│kes authen│ticated-blake2 fast│er than none and au│thenticated. │
764 │ │ │ │ │
765 On│modern AR│M CPUs, NEON provid│es hardware acceler│ation for SHA256 ma│k‐
766 in│g it faste│r than BLAKE2b-256 │there. NEON acceler│ates AES as well. │
767 │ │ │ │ │
768 Ha│rdware acc│eleration is always│used automatically│when available. │
769 │ │ │ │ │
770 re│pokey and │keyfile use AES-CTR│-256 for encryption│and HMAC-SHA256 f│or
771 au│thenticati│on in an encrypt-t│hen-MAC (EtM) const│ruction. The chunk │ID
772 ha│sh is HMAC│-SHA256 as well (wi│th a separate key)│. These modes a│re
773 co│mpatible w│ith Borg 1.0.x. │ │ │
774 │ │ │ │ │
775 re│pokey-blak│e2 and keyfile-bl│ake2 are also aut│henticated encrypti│on
776 mo│des, but u│se BLAKE2b-256 inst│ead of HMAC-SHA256 │for authenticatio│n.
777 Th│e chunk ID│hash is a keyed BL│AKE2b-256 hash. Th│ese modes are new a│nd
778 no│t compatib│le with Borg 1.0.x.│ │ │
779 │ │ │ │ │
780 au│thenticate│d mode uses no enc│ryption, but auth│enticates reposito│ry
781 co│ntents th│rough the same HMA│C-SHA256 hash as th│e repokey and keyfi│le
782 mo│des (it us│es it as the chunk│ ID hash). The │key is stored li│ke
783 re│pokey. Th│is mode is new and │not compatible with│Borg 1.0.x. │
784 │ │ │ │ │
785 au│thenticate│d-blake2 is lik│e authenticated, │but uses the key│ed
786 BL│AKE2b-256 │hash from the other│blake2 modes. Thi│s mode is new and n│ot
787 co│mpatible w│ith Borg 1.0.x. │ │ │
788 │ │ │ │ │
789 no│ne mode │uses no encryption │and no authenticati│on. It uses SHA256 │as
790 ch│unk ID has│h. Not recommended,│rather consider us│ing an authenticat│ed
791 or│ authenti│cated/encrypted mod│e. This mode has po│ssible denial-of-se│r‐
792 vi│ce issues │when running borg │create on content│s controlled by │an
793 at│tacker. │Use it only for │new repositories w│here no encryption │is
794 wa│nted and w│hen compatibility w│ith 1.0.x is import│ant. If compatibili│ty
795 wi│th 1.0.x │is not important, u│se authenticated-bl│ake2 or authenticat│ed
796 in│stead. Th│is mode is compatib│le with Borg 1.0.x.│ │
797 │ │ │ │ │
798 Exampl│es │ │ │ │
799 │# Local r│epository, repokey │encryption, BLAKE2b│(often faster, sin│ce Borg 1.1)
800 │$ borg in│it --encryption=rep│okey-blake2 /path/t│o/repo │
801 │ │ │ │ │
802 │# Local r│epository (no encry│ption) │ │
803 │$ borg in│it --encryption=non│e /path/to/repo │ │
804 │ │ │ │ │
805 │# Remote │repository (accesse│s a remote borg via│ssh) │
806 │$ borg in│it --encryption=rep│okey-blake2 user@ho│stname:backup │
807 │ │ │ │ │
808 │# Remote │repository (store t│he key your home di│r) │
809 │$ borg in│it --encryption=key│file user@hostname:│backup │
810 │ │ │ │ │
812 │borg [com│mon options] create│[options] ARCHIVE │[PATH...] │
813 │ │ │ │ │
814 Descri│ption │ │ │ │
815 Th│is command│creates a backup a│rchive containing a│ll files found whi│le
816 re│cursively │traversing all pat│hs specified. Paths│are added to the a│r‐
817 ch│ive as the│y are given, that m│eans if relative pa│ths are desired, t│he
818 co│mmand has │to be run from the │correct directory. │ │
819 When giving '-' as path, borg will read data from standard input and
821 create a file 'stdin' in the created archive from that data.
822 The archive will consume almost no disk space for files or parts of
824 files that have already been stored in other archives.
825 The archive name needs to be unique. It must not end in '.checkpoint'
827 or '.checkpoint.N' (with N being a number), because these names are
828 used for checkpoints and treated in special ways.
829 In the archive name, you may use the following placeholders: {now},
831 {utcnow}, {fqdn}, {hostname}, {user} and some others.
832 Backup speed is increased by not reprocessing files that are already
834 part of existing archives and weren't modified. The detection of unmod‐
835 ified files is done by comparing multiple file metadata values with
836 previous values kept in the files cache.
837 This comparison can operate in different modes as given by
839 --files-cache:
840 · ctime,size,inode (default)
842 · mtime,size,inode (default behaviour of borg versions older than
844 1.1.0rc4)
845 · ctime,size (ignore the inode number)
847 · mtime,size (ignore the inode number)
849 · rechunk,ctime (all files are considered modified - rechunk, cache
851 ctime)
852 · rechunk,mtime (all files are considered modified - rechunk, cache
854 mtime)
855 · disabled (disable the files cache, all files considered modified -
857 rechunk)
858 inode number: better safety, but often unstable on network filesystems
860 Normally, detecting file modifications will take inode information into
862 consideration to improve the reliability of file change detection.
863 This is problematic for files located on sshfs and similar network file
864 systems which do not provide stable inode numbers, such files will
865 always be considered modified. You can use modes without inode in this
866 case to improve performance, but reliability of change detection might
867 be reduced.
868 ctime vs. mtime: safety vs. speed
870 · ctime is a rather safe way to detect changes to a file (metadata and
872 contents) as it can not be set from userspace. But, a metadata-only
873 change will already update the ctime, so there might be some unneces‐
874 sary chunking/hashing even without content changes. Some filesystems
875 do not support ctime (change time).
876 · mtime usually works and only updates if file contents were changed.
878 But mtime can be arbitrarily set from userspace, e.g. to set mtime
879 back to the same value it had before a content change happened. This
880 can be used maliciously as well as well-meant, but in both cases
881 mtime based cache modes can be problematic.
882 The mount points of filesystems or filesystem snapshots should be the
884 same for every creation of a new archive to ensure fast operation. This
885 is because the file cache that is used to determine changed files
886 quickly uses absolute filenames. If this is not possible, consider
887 creating a bind mount to a stable location.
888 The --progress option shows (from left to right) Original, Compressed
890 and Deduplicated (O, C and D, respectively), then the Number of files
891 (N) processed so far, followed by the currently processed path.
892 When using --stats, you will get some statistics about how much data
894 was added - the "This Archive" deduplicated size there is most inter‐
895 esting as that is how much your repository will grow. Please note that
896 the "All archives" stats refer to the state after creation. Also, the
897 --stats and --dry-run options are mutually exclusive because the data
898 is not actually compressed and deduplicated during a dry run.
899 See the output of the "borg help patterns" command for more help on
901 exclude patterns. See the output of the "borg help placeholders" com‐
902 mand for more help on placeholders.
903 The --exclude patterns are not like tar. In tar --exclude .bundler/gems
905 will exclude foo/.bundler/gems. In borg it will not, you need to use
906 --exclude '*/.bundler/gems' to get the same effect. See borg help pat‐
907 terns for more information.
908 In addition to using --exclude patterns, it is possible to use
910 --exclude-if-present to specify the name of a filesystem object (e.g. a
911 file or folder name) which, when contained within another folder, will
912 prevent the containing folder from being backed up. By default, the
913 containing folder and all of its contents will be omitted from the
914 backup. If, however, you wish to only include the objects specified by
915 --exclude-if-present in your backup, and not include any other contents
916 of the containing folder, this can be enabled through using the
917 --keep-exclude-tags option.
918 Item flags
920 --list outputs a list of all files, directories and other file system
921 items it considered (no matter whether they had content changes or
922 not). For each item, it prefixes a single-letter flag that indicates
923 type and/or status of the item.
924 If you are interested only in a subset of that output, you can give
926 e.g. --filter=AME and it will only show regular files with A, M or E
927 status (see below).
928 A uppercase character represents the status of a regular file relative
930 to the "files" cache (not relative to the repo -- this is an issue if
931 the files cache is not used). Metadata is stored in any case and for
932 'A' and 'M' also new data chunks are stored. For 'U' all data chunks
933 refer to already existing chunks.
934 · 'A' = regular file, added (see also a_status_oddity in the FAQ)
936 · 'M' = regular file, modified
938 · 'U' = regular file, unchanged
940 · 'E' = regular file, an error happened while accessing/reading this
942 file
943 A lowercase character means a file type other than a regular file, borg
945 usually just stores their metadata:
946 · 'd' = directory
948 · 'b' = block device
950 · 'c' = char device
952 · 'h' = regular file, hardlink (to already seen inodes)
954 · 's' = symlink
956 · 'f' = fifo
958 Other flags used include:
960 · 'i' = backup data was read from standard input (stdin)
962 · '-' = dry run, item was not backed up
964 · 'x' = excluded, item was not backed up
966 · '?' = missing status code (if you see this, please file a bug
968 report!)
969 Examples
971 # Backup ~/Documents into an archive named "my-documents"
972 $ borg create /path/to/repo::my-documents ~/Documents
973 # same, but list all files as we process them
975 $ borg create --list /path/to/repo::my-documents ~/Documents
976 # Backup ~/Documents and ~/src but exclude pyc files
978 $ borg create /path/to/repo::my-files \
979 ~/Documents \
980 ~/src \
981 --exclude '*.pyc'
982 # Backup home directories excluding image thumbnails (i.e. only
984 # /home/<one directory>/.thumbnails is excluded, not /home/*/*/.thumbnails etc.)
985 $ borg create /path/to/repo::my-files /home \
986 --exclude 'sh:/home/*/.thumbnails'
987 # Backup the root filesystem into an archive named "root-YYYY-MM-DD"
989 # use zlib compression (good, but slow) - default is lz4 (fast, low compression ratio)
990 $ borg create -C zlib,6 --one-file-system /path/to/repo::root-{now:%Y-%m-%d} /
991 # Backup a remote host locally ("pull" style) using sshfs
993 $ mkdir sshfs-mount
994 $ sshfs root@example.com:/ sshfs-mount
995 $ cd sshfs-mount
996 $ borg create /path/to/repo::example.com-root-{now:%Y-%m-%d} .
997 $ cd ..
998 $ fusermount -u sshfs-mount
999 # Make a big effort in fine granular deduplication (big chunk management
1001 # overhead, needs a lot of RAM and disk space, see formula in internals
1002 # docs - same parameters as borg < 1.0 or attic):
1003 $ borg create --chunker-params 10,23,16,4095 /path/to/repo::small /smallstuff
1004 # Backup a raw device (must not be active/in use/mounted at that time)
1006 $ dd if=/dev/sdx bs=10M | borg create /path/to/repo::my-sdx -
1007 # No compression (none)
1009 $ borg create --compression none /path/to/repo::arch ~
1010 # Super fast, low compression (lz4, default)
1012 $ borg create /path/to/repo::arch ~
1013 # Less fast, higher compression (zlib, N = 0..9)
1015 $ borg create --compression zlib,N /path/to/repo::arch ~
1016 # Even slower, even higher compression (lzma, N = 0..9)
1018 $ borg create --compression lzma,N /path/to/repo::arch ~
1019 # Only compress compressible data with lzma,N (N = 0..9)
1021 $ borg create --compression auto,lzma,N /path/to/repo::arch ~
1022 # Use short hostname, user name and current time in archive name
1024 $ borg create /path/to/repo::{hostname}-{user}-{now} ~
1025 # Similar, use the same datetime format that is default as of borg 1.1
1026 $ borg create /path/to/repo::{hostname}-{user}-{now:%Y-%m-%dT%H:%M:%S} ~
1027 # As above, but add nanoseconds
1028 $ borg create /path/to/repo::{hostname}-{user}-{now:%Y-%m-%dT%H:%M:%S.%f} ~
1029 # Backing up relative paths by moving into the correct directory first
1031 $ cd /home/user/Documents
1032 # The root directory of the archive will be "projectA"
1033 $ borg create /path/to/repo::daily-projectA-{now:%Y-%m-%d} projectA
1034
1036 borg [common options] extract [options] ARCHIVE [PATH...]
1037 Description
1039 This command extracts the contents of an archive. By default the entire
1040 archive is extracted but a subset of files and directories can be
1041 selected by passing a list of PATHs as arguments. The file selection
1042 can further be restricted by using the --exclude option.
1043 See the output of the "borg help patterns" command for more help on
1045 exclude patterns.
1046 By using --dry-run, you can do all extraction steps except actually
1048 writing the output data: reading metadata and data chunks from the
1049 repo, checking the hash/hmac, decrypting, decompressing.
1050 --progress can be slower than no progress display, since it makes one
1052 additional pass over the archive metadata.
1053 NOTE:
1055 Currently, extract always writes into the current working directory
1056 ("."), so make sure you cd to the right place before calling borg
1057 extract.
1058 Examples
1060 # Extract entire archive
1061 $ borg extract /path/to/repo::my-files
1062 # Extract entire archive and list files while processing
1064 $ borg extract --list /path/to/repo::my-files
1065 # Verify whether an archive could be successfully extracted, but do not write files to disk
1067 $ borg extract --dry-run /path/to/repo::my-files
1068 # Extract the "src" directory
1070 $ borg extract /path/to/repo::my-files home/USERNAME/src
1071 # Extract the "src" directory but exclude object files
1073 $ borg extract /path/to/repo::my-files home/USERNAME/src --exclude '*.o'
1074 # Restore a raw device (must not be active/in use/mounted at that time)
1076 $ borg extract --stdout /path/to/repo::my-sdx | dd of=/dev/sdx bs=10M
1077
1079 borg [common options] check [options] [REPOSITORY_OR_ARCHIVE]
1080 Description
1082 The check command verifies the consistency of a repository and the cor‐
1083 responding archives.
1084 First, the underlying repository data files are checked:
1086 · For all segments the segment magic (header) is checked
1088 · For all objects stored in the segments, all metadata (e.g. crc and
1090 size) and all data is read. The read data is checked by size and CRC.
1091 Bit rot and other types of accidental damage can be detected this
1092 way.
1093 · If we are in repair mode and a integrity error is detected for a seg‐
1095 ment, we try to recover as many objects from the segment as possible.
1096 · In repair mode, it makes sure that the index is consistent with the
1098 data stored in the segments.
1099 · If you use a remote repo server via ssh:, the repo check is executed
1101 on the repo server without causing significant network traffic.
1102 · The repository check can be skipped using the --archives-only option.
1104 Second, the consistency and correctness of the archive metadata is ver‐
1106 ified:
1107 · Is the repo manifest present? If not, it is rebuilt from archive
1109 metadata chunks (this requires reading and decrypting of all metadata
1110 and data).
1111 · Check if archive metadata chunk is present. if not, remove archive
1113 from manifest.
1114 · For all files (items) in the archive, for all chunks referenced by
1116 these files, check if chunk is present. If a chunk is not present
1117 and we are in repair mode, replace it with a same-size replacement
1118 chunk of zeros. If a previously lost chunk reappears (e.g. via a
1119 later backup) and we are in repair mode, the all-zero replacement
1120 chunk will be replaced by the correct chunk. This requires reading
1121 of archive and file metadata, but not data.
1122 · If we are in repair mode and we checked all the archives: delete
1124 orphaned chunks from the repo.
1125 · if you use a remote repo server via ssh:, the archive check is exe‐
1127 cuted on the client machine (because if encryption is enabled, the
1128 checks will require decryption and this is always done client-side,
1129 because key access will be required).
1130 · The archive checks can be time consuming, they can be skipped using
1132 the --repository-only option.
1133 The --verify-data option will perform a full integrity verification (as
1135 opposed to checking the CRC32 of the segment) of data, which means
1136 reading the data from the repository, decrypting and decompressing it.
1137 This is a cryptographic verification, which will detect (accidental)
1138 corruption. For encrypted repositories it is tamper-resistant as well,
1139 unless the attacker has access to the keys.
1140 It is also very slow.
1142
1144 borg [common options] rename [options] ARCHIVE NEWNAME
1145 Description
1147 This command renames an archive in the repository.
1148 This results in a different archive ID.
1150 Examples
1152 $ borg create /path/to/repo::archivename ~
1153 $ borg list /path/to/repo
1154 archivename Mon, 2016-02-15 19:50:19
1155 $ borg rename /path/to/repo::archivename newname
1157 $ borg list /path/to/repo
1158 newname Mon, 2016-02-15 19:50:19
1159
1161 borg [common options] list [options] [REPOSITORY_OR_ARCHIVE] [PATH...]
1162 Description
1164 This command lists the contents of a repository or an archive.
1165 See the "borg help patterns" command for more help on exclude patterns.
1167 The following keys are available for --format:
1169 · NEWLINE: OS dependent line separator
1171 · NL: alias of NEWLINE
1173 · NUL: NUL character for creating print0 / xargs -0 like output, see
1175 barchive/bpath
1176 · SPACE
1178 · TAB
1180 · CR
1182 · LF
1184 Keys for listing repository archives:
1186 · archive: archive name interpreted as text (might be missing non-text
1188 characters, see barchive)
1189 · name: alias of "archive"
1191 · barchive: verbatim archive name, can contain any character except NUL
1193 · comment: archive comment interpreted as text (might be missing
1195 non-text characters, see bcomment)
1196 · bcomment: verbatim archive comment, can contain any character except
1198 NUL
1199 · id: internal ID of the archive
1201 · start: time (start) of creation of the archive
1203 · time: alias of "start"
1205 · end: time (end) of creation of the archive
1207 · hostname: hostname of host on which this archive was created
1209 · username: username of user who created this archive
1211 Keys for listing archive files:
1213 · type
1215 · mode
1217 · uid
1219 · gid
1221 · user
1223 · group
1225 · path: path interpreted as text (might be missing non-text characters,
1227 see bpath)
1228 · bpath: verbatim POSIX path, can contain any character except NUL
1230 · source: link target for links (identical to linktarget)
1232 · linktarget
1234 · flags
1236 · size
1238 · csize: compressed size
1240 · dsize: deduplicated size
1242 · dcsize: deduplicated compressed size
1244 · num_chunks: number of chunks in this file
1246 · unique_chunks: number of unique chunks in this file
1248 · mtime
1250 · ctime
1252 · atime
1254 · isomtime
1256 · isoctime
1258 · isoatime
1260 · md5
1262 · sha1
1264 · sha224
1266 · sha256
1268 · sha384
1270 · sha512
1272 · archiveid
1274 · archivename
1276 · extra: prepends {source} with " -> " for soft links and " link to "
1278 for hard links
1279 · health: either "healthy" (file ok) or "broken" (if file has all-zero
1281 replacement chunks)
1282 Examples
1284 $ borg list /path/to/repo
1285 Monday Mon, 2016-02-15 19:15:11
1286 repo Mon, 2016-02-15 19:26:54
1287 root-2016-02-15 Mon, 2016-02-15 19:36:29
1288 newname Mon, 2016-02-15 19:50:19
1289 ...
1290 $ borg list /path/to/repo::root-2016-02-15
1292 drwxr-xr-x root root 0 Mon, 2016-02-15 17:44:27 .
1293 drwxrwxr-x root root 0 Mon, 2016-02-15 19:04:49 bin
1294 -rwxr-xr-x root root 1029624 Thu, 2014-11-13 00:08:51 bin/bash
1295 lrwxrwxrwx root root 0 Fri, 2015-03-27 20:24:26 bin/bzcmp -> bzdiff
1296 -rwxr-xr-x root root 2140 Fri, 2015-03-27 20:24:22 bin/bzdiff
1297 ...
1298 $ borg list /path/to/repo::archiveA --format="{mode} {user:6} {group:6} {size:8d} {isomtime} {path}{extra}{NEWLINE}"
1300 drwxrwxr-x user user 0 Sun, 2015-02-01 11:00:00 .
1301 drwxrwxr-x user user 0 Sun, 2015-02-01 11:00:00 code
1302 drwxrwxr-x user user 0 Sun, 2015-02-01 11:00:00 code/myproject
1303 -rw-rw-r-- user user 1416192 Sun, 2015-02-01 11:00:00 code/myproject/file.ext
1304 ...
1305
1307 borg [common options] diff [options] REPO::ARCHIVE1 ARCHIVE2 [PATH...]
1308 Description
1310 This command finds differences (file contents, user/group/mode) between
1311 archives.
1312 A repository location and an archive name must be specified for
1314 REPO::ARCHIVE1. ARCHIVE2 is just another archive name in same reposi‐
1315 tory (no repository location allowed).
1316 For archives created with Borg 1.1 or newer diff automatically detects
1318 whether the archives are created with the same chunker params. If so,
1319 only chunk IDs are compared, which is very fast.
1320 For archives prior to Borg 1.1 chunk contents are compared by default.
1322 If you did not create the archives with different chunker params, pass
1323 --same-chunker-params. Note that the chunker params changed from Borg
1324 0.xx to 1.0.
1325 See the output of the "borg help patterns" command for more help on
1327 exclude patterns.
1328 Examples
1330 $ borg init -e=none testrepo
1331 $ mkdir testdir
1332 $ cd testdir
1333 $ echo asdf > file1
1334 $ dd if=/dev/urandom bs=1M count=4 > file2
1335 $ touch file3
1336 $ borg create ../testrepo::archive1 .
1337 $ chmod a+x file1
1339 $ echo "something" >> file2
1340 $ borg create ../testrepo::archive2 .
1341 $ rm file3
1343 $ touch file4
1344 $ borg create ../testrepo::archive3 .
1345 $ cd ..
1347 $ borg diff testrepo::archive1 archive2
1348 [-rw-r--r-- -> -rwxr-xr-x] file1
1349 +135 B -252 B file2
1350 $ borg diff testrepo::archive2 archive3
1352 added 0 B file4
1353 removed 0 B file3
1354 $ borg diff testrepo::archive1 archive3
1356 [-rw-r--r-- -> -rwxr-xr-x] file1
1357 +135 B -252 B file2
1358 added 0 B file4
1359 removed 0 B file3
1360
1362 borg [common options] delete [options] [REPOSITORY_OR_ARCHIVE] [ARCHIVE...]
1363 Description
1365 This command deletes an archive from the repository or the complete
1366 repository. Disk space is reclaimed accordingly. If you delete the
1367 complete repository, the local cache for it (if any) is also deleted.
1368 When using --stats, you will get some statistics about how much data
1370 was deleted - the "Deleted data" deduplicated size there is most inter‐
1371 esting as that is how much your repository will shrink. Please note
1372 that the "All archives" stats refer to the state after deletion.
1373 Examples
1375 # delete a single backup archive:
1376 $ borg delete /path/to/repo::Monday
1377 # delete the whole repository and the related local cache:
1379 $ borg delete /path/to/repo
1380 You requested to completely DELETE the repository *including* all archives it contains:
1381 repo Mon, 2016-02-15 19:26:54
1382 root-2016-02-15 Mon, 2016-02-15 19:36:29
1383 newname Mon, 2016-02-15 19:50:19
1384 Type 'YES' if you understand this and want to continue: YES
1385
1387 borg [common options] prune [options] [REPOSITORY]
1388 Description
1390 The prune command prunes a repository by deleting all archives not
1391 matching any of the specified retention options. This command is nor‐
1392 mally used by automated backup scripts wanting to keep a certain number
1393 of historic backups.
1394 Also, prune automatically removes checkpoint archives (incomplete ar‐
1396 chives left behind by interrupted backup runs) except if the checkpoint
1397 is the latest archive (and thus still needed). Checkpoint archives are
1398 not considered when comparing archive counts against the retention lim‐
1399 its (--keep-X).
1400 If a prefix is set with -P, then only archives that start with the pre‐
1402 fix are considered for deletion and only those archives count towards
1403 the totals specified by the rules. Otherwise, all archives in the
1404 repository are candidates for deletion! There is no automatic distinc‐
1405 tion between archives representing different contents. These need to be
1406 distinguished by specifying matching prefixes.
1407 If you have multiple sequences of archives with different data sets
1409 (e.g. from different machines) in one shared repository, use one prune
1410 call per data set that matches only the respective archives using the
1411 -P option.
1412 The --keep-within option takes an argument of the form "<int><char>",
1414 where char is "H", "d", "w", "m", "y". For example, --keep-within 2d
1415 means to keep all archives that were created within the past 48 hours.
1416 "1m" is taken to mean "31d". The archives kept with this option do not
1417 count towards the totals specified by any other options.
1418 A good procedure is to thin out more and more the older your backups
1420 get. As an example, --keep-daily 7 means to keep the latest backup on
1421 each day, up to 7 most recent days with backups (days without backups
1422 do not count). The rules are applied from secondly to yearly, and
1423 backups selected by previous rules do not count towards those of later
1424 rules. The time that each backup starts is used for pruning purposes.
1425 Dates and times are interpreted in the local timezone, and weeks go
1426 from Monday to Sunday. Specifying a negative number of archives to keep
1427 means that there is no limit.
1428 The --keep-last N option is doing the same as --keep-secondly N (and it
1430 will keep the last N archives under the assumption that you do not cre‐
1431 ate more than one backup archive in the same second).
1432 When using --stats, you will get some statistics about how much data
1434 was deleted - the "Deleted data" deduplicated size there is most inter‐
1435 esting as that is how much your repository will shrink. Please note
1436 that the "All archives" stats refer to the state after pruning.
1437 Examples
1439 Be careful, prune is a potentially dangerous command, it will remove
1440 backup archives.
1441 The default of prune is to apply to all archives in the repository
1443 unless you restrict its operation to a subset of the archives using
1444 --prefix. When using --prefix, be careful to choose a good prefix -
1445 e.g. do not use a prefix "foo" if you do not also want to match "foo‐
1446 bar".
1447 It is strongly recommended to always run prune -v --list --dry-run ...
1449 first so you will see what it would do without it actually doing any‐
1450 thing.
1451 # Keep 7 end of day and 4 additional end of week archives.
1453 # Do a dry-run without actually deleting anything.
1454 $ borg prune -v --list --dry-run --keep-daily=7 --keep-weekly=4 /path/to/repo
1455 # Same as above but only apply to archive names starting with the hostname
1457 # of the machine followed by a "-" character:
1458 $ borg prune -v --list --keep-daily=7 --keep-weekly=4 --prefix='{hostname}-' /path/to/repo
1459 # Keep 7 end of day, 4 additional end of week archives,
1461 # and an end of month archive for every month:
1462 $ borg prune -v --list --keep-daily=7 --keep-weekly=4 --keep-monthly=-1 /path/to/repo
1463 # Keep all backups in the last 10 days, 4 additional end of week archives,
1465 # and an end of month archive for every month:
1466 $ borg prune -v --list --keep-within=10d --keep-weekly=4 --keep-monthly=-1 /path/to/repo
1467 There is also a visualized prune example in docs/misc/prune-exam‐
1469 ple.txt:
1470 borg prune visualized
1472 =====================
1473 Assume it is 2016-01-01, today's backup has not yet been made and you have
1475 created at least one backup on each day in 2015 except on 2015-12-19 (no
1476 backup made on that day).
1477 This is what borg prune --keep-daily 14 --keep-monthly 6 would keep.
1479 Backups kept by the --keep-daily rule are marked by a "d" to the right,
1481 backups kept by the --keep-monthly rule are marked by a "m" to the right.
1482 Calendar view
1484 -------------
1485 2015
1487 January February March
1488 Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su
1489 1 2 3 4 1 1
1490 5 6 7 8 9 10 11 2 3 4 5 6 7 8 2 3 4 5 6 7 8
1491 12 13 14 15 16 17 18 9 10 11 12 13 14 15 9 10 11 12 13 14 15
1492 19 20 21 22 23 24 25 16 17 18 19 20 21 22 16 17 18 19 20 21 22
1493 26 27 28 29 30 31 23 24 25 26 27 28 23 24 25 26 27 28 29
1494 30 31
1495 April May June
1497 Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su
1498 1 2 3 4 5 1 2 3 1 2 3 4 5 6 7
1499 6 7 8 9 10 11 12 4 5 6 7 8 9 10 8 9 10 11 12 13 14
1500 13 14 15 16 17 18 19 11 12 13 14 15 16 17 15 16 17 18 19 20 21
1501 20 21 22 23 24 25 26 18 19 20 21 22 23 24 22 23 24 25 26 27 28
1502 27 28 29 30 25 26 27 28 29 30 31 29 30m
1503 July August September
1506 Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su
1507 1 2 3 4 5 1 2 1 2 3 4 5 6
1508 6 7 8 9 10 11 12 3 4 5 6 7 8 9 7 8 9 10 11 12 13
1509 13 14 15 16 17 18 19 10 11 12 13 14 15 16 14 15 16 17 18 19 20
1510 20 21 22 23 24 25 26 17 18 19 20 21 22 23 21 22 23 24 25 26 27
1511 27 28 29 30 31m 24 25 26 27 28 29 30 28 29 30m
1512 31m
1513 October November December
1515 Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su
1516 1 2 3 4 1 1 2 3 4 5 6
1517 5 6 7 8 9 10 11 2 3 4 5 6 7 8 7 8 9 10 11 12 13
1518 12 13 14 15 16 17 18 9 10 11 12 13 14 15 14 15 16 17d18d19 20d
1519 19 20 21 22 23 24 25 16 17 18 19 20 21 22 21d22d23d24d25d26d27d
1520 26 27 28 29 30 31m 23 24 25 26 27 28 29 28d29d30d31d
1521 30m
1522 List view
1524 ---------
1525 --keep-daily 14 --keep-monthly 6
1527 -------------------------------------------------
1528 1. 2015-12-31 (2015-12-31 kept by daily rule)
1529 2. 2015-12-30 1. 2015-11-30
1530 3. 2015-12-29 2. 2015-10-31
1531 4. 2015-12-28 3. 2015-09-30
1532 5. 2015-12-27 4. 2015-08-31
1533 6. 2015-12-26 5. 2015-07-31
1534 7. 2015-12-25 6. 2015-06-30
1535 8. 2015-12-24
1536 9. 2015-12-23
1537 10. 2015-12-22
1538 11. 2015-12-21
1539 12. 2015-12-20
1540 (no backup made on 2015-12-19)
1541 13. 2015-12-18
1542 14. 2015-12-17
1543 Notes
1546 -----
1547 2015-12-31 is kept due to the --keep-daily 14 rule (because it is applied
1549 first), not due to the --keep-monthly rule.
1550 Because of that, the --keep-monthly 6 rule keeps Nov, Oct, Sep, Aug, Jul and
1552 Jun. December is not considered for this rule, because that backup was already
1553 kept because of the daily rule.
1554 2015-12-17 is kept to satisfy the --keep-daily 14 rule - because no backup was
1556 made on 2015-12-19. If a backup had been made on that day, it would not keep
1557 the one from 2015-12-17.
1558 We did not include yearly, weekly, hourly, minutely or secondly rules to keep
1560 this example simple. They all work in basically the same way.
1561 The weekly rule is easy to understand roughly, but hard to understand in all
1563 details. If interested, read "ISO 8601:2000 standard week-based year".
1564
1567 borg [common options] info [options] [REPOSITORY_OR_ARCHIVE]
1568 Description
1570 This command displays detailed information about the specified archive
1571 or repository.
1572 Please note that the deduplicated sizes of the individual archives do
1574 not add up to the deduplicated size of the repository ("all archives"),
1575 because the two are meaning different things:
1576 This archive / deduplicated size = amount of data stored ONLY for this
1578 archive = unique chunks of this archive. All archives / deduplicated
1579 size = amount of data stored in the repo = all chunks in the reposi‐
1580 tory.
1581 Borg archives can only contain a limited amount of file metadata. The
1583 size of an archive relative to this limit depends on a number of fac‐
1584 tors, mainly the number of files, the lengths of paths and other meta‐
1585 data stored for files. This is shown as utilization of maximum sup‐
1586 ported archive size.
1587 Examples
1589 $ borg info /path/to/repo::2017-06-29T11:00-srv
1590 Archive name: 2017-06-29T11:00-srv
1591 Archive fingerprint: b2f1beac2bd553b34e06358afa45a3c1689320d39163890c5bbbd49125f00fe5
1592 Comment:
1593 Hostname: myhostname
1594 Username: root
1595 Time (start): Thu, 2017-06-29 11:03:07
1596 Time (end): Thu, 2017-06-29 11:03:13
1597 Duration: 5.66 seconds
1598 Number of files: 17037
1599 Command line: /usr/sbin/borg create /path/to/repo::2017-06-29T11:00-srv /srv
1600 Utilization of max. archive size: 0%
1601 ------------------------------------------------------------------------------
1602 Original size Compressed size Deduplicated size
1603 This archive: 12.53 GB 12.49 GB 1.62 kB
1604 All archives: 121.82 TB 112.41 TB 215.42 GB
1605 Unique chunks Total chunks
1607 Chunk index: 1015213 626934122
1608 $ borg info /path/to/repo --last 1
1610 Archive name: 2017-06-29T11:00-srv
1611 Archive fingerprint: b2f1beac2bd553b34e06358afa45a3c1689320d39163890c5bbbd49125f00fe5
1612 Comment:
1613 Hostname: myhostname
1614 Username: root
1615 Time (start): Thu, 2017-06-29 11:03:07
1616 Time (end): Thu, 2017-06-29 11:03:13
1617 Duration: 5.66 seconds
1618 Number of files: 17037
1619 Command line: /usr/sbin/borg create /path/to/repo::2017-06-29T11:00-srv /srv
1620 Utilization of max. archive size: 0%
1621 ------------------------------------------------------------------------------
1622 Original size Compressed size Deduplicated size
1623 This archive: 12.53 GB 12.49 GB 1.62 kB
1624 All archives: 121.82 TB 112.41 TB 215.42 GB
1625 Unique chunks Total chunks
1627 Chunk index: 1015213 626934122
1628 $ borg info /path/to/repo
1630 Repository ID: d857ce5788c51272c61535062e89eac4e8ef5a884ffbe976e0af9d8765dedfa5
1631 Location: /path/to/repo
1632 Encrypted: Yes (repokey)
1633 Cache: /root/.cache/borg/d857ce5788c51272c61535062e89eac4e8ef5a884ffbe976e0af9d8765dedfa5
1634 Security dir: /root/.config/borg/security/d857ce5788c51272c61535062e89eac4e8ef5a884ffbe976e0af9d8765dedfa5
1635 ------------------------------------------------------------------------------
1636 Original size Compressed size Deduplicated size
1637 All archives: 121.82 TB 112.41 TB 215.42 GB
1638 Unique chunks Total chunks
1640 Chunk index: 1015213 626934122
1641
1643 borg [common options] mount [options] REPOSITORY_OR_ARCHIVE MOUNTPOINT [PATH...]
1644 Description
1646 This command mounts an archive as a FUSE filesystem. This can be useful
1647 for browsing an archive or restoring individual files. Unless the
1648 --foreground option is given the command will run in the background
1649 until the filesystem is umounted.
1650 The command borgfs provides a wrapper for borg mount. This can also be
1652 used in fstab entries: /path/to/repo /mnt/point fuse.borgfs
1653 defaults,noauto 0 0
1654 To allow a regular user to use fstab entries, add the user option:
1656 /path/to/repo /mnt/point fuse.borgfs defaults,noauto,user 0 0
1657 For FUSE configuration and mount options, see the mount.fuse(8) manual
1659 page.
1660 Additional mount options supported by borg:
1662 · versions: when used with a repository mount, this gives a merged,
1664 versioned view of the files in the archives. EXPERIMENTAL, layout may
1665 change in future.
1666 · allow_damaged_files: by default damaged files (where missing chunks
1668 were replaced with runs of zeros by borg check --repair) are not
1669 readable and return EIO (I/O error). Set this option to read such
1670 files.
1671 · ignore_permissions: for security reasons the "default_permissions"
1673 mount option is internally enforced by borg. "ignore_permissions" can
1674 be given to not enforce "default_permissions".
1675 The BORG_MOUNT_DATA_CACHE_ENTRIES environment variable is meant for
1677 advanced users to tweak the performance. It sets the number of cached
1678 data chunks; additional memory usage can be up to ~8 MiB times this
1679 number. The default is the number of CPU cores.
1680 When the daemonized process receives a signal or crashes, it does not
1682 unmount. Unmounting in these cases could cause an active rsync or sim‐
1683 ilar process to unintentionally delete data.
1684 When running in the foreground ^C/SIGINT unmounts cleanly, but other
1686 signals or crashes do not.
1687
1689 borg [common options] umount [options] MOUNTPOINT
1690 Description
1692 This command un-mounts a FUSE filesystem that was mounted with borg
1693 mount.
1694 This is a convenience wrapper that just calls the platform-specific
1696 shell command - usually this is either umount or fusermount -u.
1697 Examples
1699 # Mounting the repository shows all archives.
1700 # Archives are loaded lazily, expect some delay when navigating to an archive
1701 # for the first time.
1702 $ borg mount /path/to/repo /tmp/mymountpoint
1703 $ ls /tmp/mymountpoint
1704 root-2016-02-14 root-2016-02-15
1705 $ borg umount /tmp/mymountpoint
1706 # Mounting a specific archive is possible as well.
1708 $ borg mount /path/to/repo::root-2016-02-15 /tmp/mymountpoint
1709 $ ls /tmp/mymountpoint
1710 bin boot etc home lib lib64 lost+found media mnt opt
1711 root sbin srv tmp usr var
1712 $ borg umount /tmp/mymountpoint
1713 # The experimental "versions view" merges all archives in the repository
1715 # and provides a versioned view on files.
1716 $ borg mount -o versions /path/to/repo /tmp/mymountpoint
1717 $ ls -l /tmp/mymountpoint/home/user/doc.txt/
1718 total 24
1719 -rw-rw-r-- 1 user group 12357 Aug 26 21:19 doc.cda00bc9.txt
1720 -rw-rw-r-- 1 user group 12204 Aug 26 21:04 doc.fa760f28.txt
1721 $ borg umount /tmp/mymountpoint
1722 # Archive filters are supported.
1724 # These are especially handy for the "versions view",
1725 # which does not support lazy processing of archives.
1726 $ borg mount -o versions --glob-archives '*-my-home' --last 10 /path/to/repo /tmp/mymountpoint
1727 # Exclusion options are supported.
1729 # These can speed up mounting and lower memory needs significantly.
1730 $ borg mount /path/to/repo /tmp/mymountpoint only/that/path
1731 $ borg mount --exclude '...' /path/to/repo /tmp/mymountpoint
1732 borgfs
1734 $ echo '/mnt/backup /tmp/myrepo fuse.borgfs defaults,noauto 0 0' >> /etc/fstab
1735 $ echo '/mnt/backup::root-2016-02-15 /tmp/myarchive fuse.borgfs defaults,noauto 0 0' >> /etc/fstab
1736 $ mount /tmp/myrepo
1737 $ mount /tmp/myarchive
1738 $ ls /tmp/myrepo
1739 root-2016-02-01 root-2016-02-2015
1740 $ ls /tmp/myarchive
1741 bin boot etc home lib lib64 lost+found media mnt opt root sbin srv tmp usr var
1742 NOTE:
1744 borgfs will be automatically provided if you used a distribution
1745 package, pip or setup.py to install Borg. Users of the standalone
1746 binary will have to manually create a symlink (see pyin‐
1747 staller-binary).
1748
1750 borg [common options] key change-passphrase [options] [REPOSITORY]
1751 Description
1753 The key files used for repository encryption are optionally passphrase
1754 protected. This command can be used to change this passphrase.
1755 Please note that this command only changes the passphrase, but not any
1757 secret protected by it (like e.g. encryption/MAC keys or chunker seed).
1758 Thus, changing the passphrase after passphrase and borg key got compro‐
1759 mised does not protect future (nor past) backups to the same reposi‐
1760 tory.
1761 Examples
1763 # Create a key file protected repository
1764 $ borg init --encryption=keyfile -v /path/to/repo
1765 Initializing repository at "/path/to/repo"
1766 Enter new passphrase:
1767 Enter same passphrase again:
1768 Remember your passphrase. Your data will be inaccessible without it.
1769 Key in "/root/.config/borg/keys/mnt_backup" created.
1770 Keep this key safe. Your data will be inaccessible without it.
1771 Synchronizing chunks cache...
1772 Archives: 0, w/ cached Idx: 0, w/ outdated Idx: 0, w/o cached Idx: 0.
1773 Done.
1774 # Change key file passphrase
1776 $ borg key change-passphrase -v /path/to/repo
1777 Enter passphrase for key /root/.config/borg/keys/mnt_backup:
1778 Enter new passphrase:
1779 Enter same passphrase again:
1780 Remember your passphrase. Your data will be inaccessible without it.
1781 Key updated
1782 Fully automated using environment variables:
1784 $ BORG_NEW_PASSPHRASE=old borg init -e=repokey repo
1786 # now "old" is the current passphrase.
1787 $ BORG_PASSPHRASE=old BORG_NEW_PASSPHRASE=new borg key change-passphrase repo
1788 # now "new" is the current passphrase.
1789
1791 borg [common options] key export [options] [REPOSITORY] [PATH]
1792 Description
1794 If repository encryption is used, the repository is inaccessible with‐
1795 out the key. This command allows to backup this essential key. Note
1796 that the backup produced does not include the passphrase itself (i.e.
1797 the exported key stays encrypted). In order to regain access to a
1798 repository, one needs both the exported key and the original
1799 passphrase.
1800 There are two backup formats. The normal backup format is suitable for
1802 digital storage as a file. The --paper backup format is optimized for
1803 printing and typing in while importing, with per line checks to reduce
1804 problems with manual input.
1805 For repositories using keyfile encryption the key is saved locally on
1807 the system that is capable of doing backups. To guard against loss of
1808 this key, the key needs to be backed up independently of the main data
1809 backup.
1810 For repositories using the repokey encryption the key is saved in the
1812 repository in the config file. A backup is thus not strictly needed,
1813 but guards against the repository becoming inaccessible if the file is
1814 damaged for some reason.
1815
1817 borg [common options] key import [options] [REPOSITORY] [PATH]
1818 Description
1820 This command allows to restore a key previously backed up with the
1821 export command.
1822 If the --paper option is given, the import will be an interactive
1824 process in which each line is checked for plausibility before proceed‐
1825 ing to the next line. For this format PATH must not be given.
1826
1828 borg [common options] upgrade [options] [REPOSITORY]
1829 Description
1831 Upgrade an existing, local Borg repository.
1832 When you do not need borg upgrade
1834 Not every change requires that you run borg upgrade.
1835 You do not need to run it when:
1837 · moving your repository to a different place
1839 · upgrading to another point release (like 1.0.x to 1.0.y), except when
1841 noted otherwise in the changelog
1842 · upgrading from 1.0.x to 1.1.x, except when noted otherwise in the
1844 changelog
1845 Borg 1.x.y upgrades
1847 Use borg upgrade --tam REPO to require manifest authentication intro‐
1848 duced with Borg 1.0.9 to address security issues. This means that modi‐
1849 fying the repository after doing this with a version prior to 1.0.9
1850 will raise a validation error, so only perform this upgrade after
1851 updating all clients using the repository to 1.0.9 or newer.
1852 This upgrade should be done on each client for safety reasons.
1854 If a repository is accidentally modified with a pre-1.0.9 client after
1856 this upgrade, use borg upgrade --tam --force REPO to remedy it.
1857 If you routinely do this you might not want to enable this upgrade
1859 (which will leave you exposed to the security issue). You can reverse
1860 the upgrade by issuing borg upgrade --disable-tam REPO.
1861 See
1863 https://borgbackup.readthedocs.io/en/stable/changes.html#pre-1-0-9-manifest-spoofing-vulnerability
1864 for details.
1865 Attic and Borg 0.xx to Borg 1.x
1867 This currently supports converting an Attic repository to Borg and also
1868 helps with converting Borg 0.xx to 1.0.
1869 Currently, only LOCAL repositories can be upgraded (issue #465).
1871 Please note that borg create (since 1.0.0) uses bigger chunks by
1873 default than old borg or attic did, so the new chunks won't deduplicate
1874 with the old chunks in the upgraded repository. See --chunker-params
1875 option of borg create and borg recreate.
1876 borg upgrade will change the magic strings in the repository's segments
1878 to match the new Borg magic strings. The keyfiles found in
1879 $ATTIC_KEYS_DIR or ~/.attic/keys/ will also be converted and copied to
1880 $BORG_KEYS_DIR or ~/.config/borg/keys.
1881 The cache files are converted, from $ATTIC_CACHE_DIR or ~/.cache/attic
1883 to $BORG_CACHE_DIR or ~/.cache/borg, but the cache layout between Borg
1884 and Attic changed, so it is possible the first backup after the conver‐
1885 sion takes longer than expected due to the cache resync.
1886 Upgrade should be able to resume if interrupted, although it will still
1888 iterate over all segments. If you want to start from scratch, use borg
1889 delete over the copied repository to make sure the cache files are also
1890 removed:
1891 borg delete borg
1893 Unless --inplace is specified, the upgrade process first creates a
1895 backup copy of the repository, in REPOSITORY.before-upgrade-DATETIME,
1896 using hardlinks. This requires that the repository and its parent
1897 directory reside on same filesystem so the hardlink copy can work.
1898 This takes longer than in place upgrades, but is much safer and gives
1899 progress information (as opposed to cp -al). Once you are satisfied
1900 with the conversion, you can safely destroy the backup copy.
1901 WARNING: Running the upgrade in place will make the current copy unus‐
1903 able with older version, with no way of going back to previous ver‐
1904 sions. This can PERMANENTLY DAMAGE YOUR REPOSITORY! Attic CAN NOT READ
1905 BORG REPOSITORIES, as the magic strings have changed. You have been
1906 warned.
1907 Examples
1909 # Upgrade the borg repository to the most recent version.
1910 $ borg upgrade -v /path/to/repo
1911 making a hardlink copy in /path/to/repo.before-upgrade-2016-02-15-20:51:55
1912 opening attic repository with borg and converting
1913 no key file found for repository
1914 converting repo index /path/to/repo/index.0
1915 converting 1 segments...
1916 converting borg 0.xx to borg current
1917 no key file found for repository
1918 Upgrading a passphrase encrypted attic repo
1920 attic offered a "passphrase" encryption mode, but this was removed in
1921 borg 1.0 and replaced by the "repokey" mode (which stores the
1922 passphrase-protected encryption key into the repository config).
1923 Thus, to upgrade a "passphrase" attic repo to a "repokey" borg repo, 2
1925 steps are needed, in this order:
1926 · borg upgrade repo
1928 · borg key migrate-to-repokey repo
1930
1932 borg [common options] recreate [options] [REPOSITORY_OR_ARCHIVE] [PATH...]
1933 Description
1935 Recreate the contents of existing archives.
1936 This is an experimental feature. Do not use this on your only backup.
1938 --exclude, --exclude-from, --exclude-if-present, --keep-exclude-tags,
1940 and PATH have the exact same semantics as in "borg create". If PATHs
1941 are specified the resulting archive will only contain files from these
1942 PATHs.
1943 Note that all paths in an archive are relative, therefore absolute pat‐
1945 terns/paths will not match (--exclude, --exclude-from, PATHs).
1946 --recompress allows to change the compression of existing data in ar‐
1948 chives. Due to how Borg stores compressed size information this might
1949 display incorrect information for archives that were not recreated at
1950 the same time. There is no risk of data loss by this.
1951 --chunker-params will re-chunk all files in the archive, this can be
1953 used to have upgraded Borg 0.xx or Attic archives deduplicate with Borg
1954 1.x archives.
1955 USE WITH CAUTION. Depending on the PATHs and patterns given, recreate
1957 can be used to permanently delete files from archives. When in doubt,
1958 use --dry-run --verbose --list to see how patterns/PATHS are inter‐
1959 preted.
1960 The archive being recreated is only removed after the operation com‐
1962 pletes. The archive that is built during the operation exists at the
1963 same time at "<ARCHIVE>.recreate". The new archive will have a differ‐
1964 ent archive ID.
1965 With --target the original archive is not replaced, instead a new ar‐
1967 chive is created.
1968 When rechunking space usage can be substantial, expect at least the
1970 entire deduplicated size of the archives using the previous chunker
1971 params. When recompressing expect approx. (throughput / check‐
1972 point-interval) in space usage, assuming all chunks are recompressed.
1973 If you recently ran borg check --repair and it had to fix lost chunks
1975 with all-zero replacement chunks, please first run another backup for
1976 the same data and re-run borg check --repair afterwards to heal any ar‐
1977 chives that had lost chunks which are still generated from the input
1978 data.
1979 Important: running borg recreate to re-chunk will remove the
1981 chunks_healthy metadata of all items with replacement chunks, so heal‐
1982 ing will not be possible any more after re-chunking (it is also
1983 unlikely it would ever work: due to the change of chunking parameters,
1984 the missing chunk likely will never be seen again even if you still
1985 have the data that produced it).
1986 Examples
1988 # Make old (Attic / Borg 0.xx) archives deduplicate with Borg 1.x archives.
1989 # Archives created with Borg 1.1+ and the default chunker params are skipped
1990 # (archive ID stays the same).
1991 $ borg recreate /mnt/backup --chunker-params default --progress
1992 # Create a backup with little but fast compression
1994 $ borg create /mnt/backup::archive /some/files --compression lz4
1995 # Then compress it - this might take longer, but the backup has already completed,
1996 # so no inconsistencies from a long-running backup job.
1997 $ borg recreate /mnt/backup::archive --recompress --compression zlib,9
1998 # Remove unwanted files from all archives in a repository.
2000 # Note the relative path for the --exclude option - archives only contain relative paths.
2001 $ borg recreate /mnt/backup --exclude home/icke/Pictures/drunk_photos
2002 # Change archive comment
2004 $ borg create --comment "This is a comment" /mnt/backup::archivename ~
2005 $ borg info /mnt/backup::archivename
2006 Name: archivename
2007 Fingerprint: ...
2008 Comment: This is a comment
2009 ...
2010 $ borg recreate --comment "This is a better comment" /mnt/backup::archivename
2011 $ borg info /mnt/backup::archivename
2012 Name: archivename
2013 Fingerprint: ...
2014 Comment: This is a better comment
2015 ...
2016
2018 borg [common options] export-tar [options] ARCHIVE FILE [PATH...]
2019 Description
2021 This command creates a tarball from an archive.
2022 When giving '-' as the output FILE, Borg will write a tar stream to
2024 standard output.
2025 By default (--tar-filter=auto) Borg will detect whether the FILE should
2027 be compressed based on its file extension and pipe the tarball through
2028 an appropriate filter before writing it to FILE:
2029 · .tar.gz: gzip
2031 · .tar.bz2: bzip2
2033 · .tar.xz: xz
2035 Alternatively a --tar-filter program may be explicitly specified. It
2037 should read the uncompressed tar stream from stdin and write a com‐
2038 pressed/filtered tar stream to stdout.
2039 The generated tarball uses the GNU tar format.
2041 export-tar is a lossy conversion: BSD flags, ACLs, extended attributes
2043 (xattrs), atime and ctime are not exported. Timestamp resolution is
2044 limited to whole seconds, not the nanosecond resolution otherwise sup‐
2045 ported by Borg.
2046 A --sparse option (as found in borg extract) is not supported.
2048 By default the entire archive is extracted but a subset of files and
2050 directories can be selected by passing a list of PATHs as arguments.
2051 The file selection can further be restricted by using the --exclude
2052 option.
2053 See the output of the "borg help patterns" command for more help on
2055 exclude patterns.
2056 --progress can be slower than no progress display, since it makes one
2058 additional pass over the archive metadata.
2059 Examples
2061 # export as uncompressed tar
2062 $ borg export-tar /path/to/repo::Monday Monday.tar
2063 # exclude some types, compress using gzip
2065 $ borg export-tar /path/to/repo::Monday Monday.tar.gz --exclude '*.so'
2066 # use higher compression level with gzip
2068 $ borg export-tar --tar-filter="gzip -9" testrepo::linux Monday.tar.gz
2069 # export a tar, but instead of storing it on disk,
2071 # upload it to a remote site using curl.
2072 $ borg export-tar /path/to/repo::Monday - | curl --data-binary @- https://somewhere/to/POST
2073 # remote extraction via "tarpipe"
2075 $ borg export-tar /path/to/repo::Monday - | ssh somewhere "cd extracted; tar x"
2076
2078 borg [common options] serve [options]
2079 Description
2081 This command starts a repository server process. This command is usu‐
2082 ally not used manually.
2083 Examples
2085 borg serve has special support for ssh forced commands (see autho‐
2086 rized_keys example below): it will detect that you use such a forced
2087 command and extract the value of the --restrict-to-path option(s).
2088 It will then parse the original command that came from the client,
2090 makes sure that it is also borg serve and enforce path restriction(s)
2091 as given by the forced command. That way, other options given by the
2092 client (like --info or --umask) are preserved (and are not fixed by the
2093 forced command).
2094 Environment variables (such as BORG_HOSTNAME_IS_UNIQUE) contained in
2096 the original command sent by the client are not interpreted, but
2097 ignored. If BORG_XXX environment variables should be set on the borg
2098 serve side, then these must be set in system-specific locations like
2099 /etc/environment or in the forced command itself (example below).
2100 # Allow an SSH keypair to only run borg, and only have access to /path/to/repo.
2102 # Use key options to disable unneeded and potentially dangerous SSH functionality.
2103 # This will help to secure an automated remote backup system.
2104 $ cat ~/.ssh/authorized_keys
2105 command="borg serve --restrict-to-path /path/to/repo",restrict ssh-rsa AAAAB3[...]
2106 # Set a BORG_XXX environment variable on the "borg serve" side
2108 $ cat ~/.ssh/authorized_keys
2109 command="export BORG_XXX=value; borg serve [...]",restrict ssh-rsa [...]
2110 NOTE:
2112 The examples above use the restrict directive. This does automati‐
2113 cally block potential dangerous ssh features, even when they are
2114 added in a future update. Thus, this option should be preferred.
2115 If you're using openssh-server < 7.2, however, you have to explic‐
2117 itly specify the ssh features to restrict and cannot simply use the
2118 restrict option as it has been introduced in v7.2. We recommend to
2119 use no-port-forwarding,no-X11-forwarding,no-pty,no-agent-forward‐
2120 ing,no-user-rc in this case.
2121 SSH Configuration
2123 borg serve's pipes (stdin/stdout/stderr) are connected to the sshd
2124 process on the server side. In the event that the SSH connection
2125 between borg serve and the client is disconnected or stuck abnormally
2126 (for example, due to a network outage), it can take a long time for
2127 sshd to notice the client is disconnected. In the meantime, sshd con‐
2128 tinues running, and as a result so does the borg serve process holding
2129 the lock on the repository. This can cause subsequent borg operations
2130 on the remote repository to fail with the error: Failed to cre‐
2131 ate/acquire the lock.
2132 In order to avoid this, it is recommended to perform the following
2134 additional SSH configuration:
2135 Either in the client side's ~/.ssh/config file, or in the client's
2137 /etc/ssh/ssh_config file:
2138 Host backupserver
2140 ServerAliveInterval 10
2141 ServerAliveCountMax 30
2142 Replacing backupserver with the hostname, FQDN or IP address of the
2144 borg server.
2145 This will cause the client to send a keepalive to the server every 10
2147 seconds. If 30 consecutive keepalives are sent without a response (a
2148 time of 300 seconds), the ssh client process will be terminated, caus‐
2149 ing the borg process to terminate gracefully.
2150 On the server side's sshd configuration file (typically
2152 /etc/ssh/sshd_config):
2153 ClientAliveInterval 10
2155 ClientAliveCountMax 30
2156 This will cause the server to send a keep alive to the client every 10
2158 seconds. If 30 consecutive keepalives are sent without a response (a
2159 time of 300 seconds), the server's sshd process will be terminated,
2160 causing the borg serve process to terminate gracefully and release the
2161 lock on the repository.
2162 If you then run borg commands with --lock-wait 600, this gives suffi‐
2164 cient time for the borg serve processes to terminate after the SSH con‐
2165 nection is torn down after the 300 second wait for the keepalives to
2166 fail.
2167 You may, of course, modify the timeout values demonstrated above to
2169 values that suit your environment and use case.
2170
2172 borg [common options] config [options] [REPOSITORY] [NAME] [VALUE]
2173 Description
2175 This command gets and sets options in a local repository or cache con‐
2176 fig file. For security reasons, this command only works on local
2177 repositories.
2178 To delete a config value entirely, use --delete. To list the values of
2180 the configuration file or the default values, use --list. To get and
2181 existing key, pass only the key name. To set a key, pass both the key
2182 name and the new value. Keys can be specified in the format "sec‐
2183 tion.name" or simply "name"; the section will default to "repository"
2184 and "cache" for the repo and cache configs, respectively.
2185 By default, borg config manipulates the repository config file. Using
2187 --cache edits the repository cache's config file instead.
2188 NOTE:
2190 The repository & cache config files are some of the only directly
2191 manipulable parts of a repository that aren't versioned or backed
2192 up, so be careful when making changes!
2193 Examples
2195 # find cache directory
2196 $ cd ~/.cache/borg/$(borg config /path/to/repo id)
2197 # reserve some space
2199 $ borg config /path/to/repo additional_free_space 2G
2200 # make a repo append-only
2202 $ borg config /path/to/repo append_only 1
2203
2205 borg [common options] with-lock [options] REPOSITORY COMMAND [ARGS...]
2206 Description
2208 This command runs a user-specified command while the repository lock is
2209 held.
2210 It will first try to acquire the lock (make sure that no other opera‐
2212 tion is running in the repo), then execute the given command as a sub‐
2213 process and wait for its termination, release the lock and return the
2214 user command's return code as borg's return code.
2215 NOTE:
2217 If you copy a repository with the lock held, the lock will be
2218 present in the copy. Thus, before using borg on the copy from a dif‐
2219 ferent host, you need to use "borg break-lock" on the copied reposi‐
2220 tory, because Borg is cautious and does not automatically remove
2221 stale locks made by a different host.
2222
2224 borg [common options] break-lock [options] [REPOSITORY]
2225 Description
2227 This command breaks the repository and cache locks. Please use care‐
2228 fully and only while no borg process (on any machine) is trying to
2229 access the Cache or the Repository.
2230
2232 borg [common options] benchmark crud [options] REPOSITORY PATH
2233 Description
2235 This command benchmarks borg CRUD (create, read, update, delete) opera‐
2236 tions.
2237 It creates input data below the given PATH and backups this data into
2239 the given REPO. The REPO must already exist (it could be a fresh empty
2240 repo or an existing repo, the command will create / read / update /
2241 delete some archives named borg-benchmark-crud* there.
2242 Make sure you have free space there, you'll need about 1GB each (+
2244 overhead).
2245 If your repository is encrypted and borg needs a passphrase to unlock
2247 the key, use:
2248 BORG_PASSPHRASE=mysecret borg benchmark crud REPO PATH
2250 Measurements are done with different input file sizes and counts. The
2252 file contents are very artificial (either all zero or all random), thus
2253 the measurement results do not necessarily reflect performance with
2254 real data. Also, due to the kind of content used, no compression is
2255 used in these benchmarks.
2256 C- == borg create (1st archive creation, no compression, do not use
2258 files cache)
2259 C-Z- == all-zero files. full dedup, this is primarily measuring
2260 reader/chunker/hasher. C-R- == random files. no dedup, measur‐
2261 ing throughput through all processing stages.
2262 R- == borg extract (extract archive, dry-run, do everything, but do not
2264 write files to disk)
2265 R-Z- == all zero files. Measuring heavily duplicated files.
2266 R-R- == random files. No duplication here, measuring throughput
2267 through all processing stages, except writing to disk.
2268 U- == borg create (2nd archive creation of unchanged input files, mea‐
2270 sure files cache speed)
2271 The throughput value is kind of virtual here, it does not actu‐
2272 ally read the file. U-Z- == needs to check the 2 all-zero
2273 chunks' existence in the repo. U-R- == needs to check existence
2274 of a lot of different chunks in the repo.
2275 D- == borg delete archive (delete last remaining archive, measure dele‐
2277 tion + compaction)
2278 D-Z- == few chunks to delete / few segments to compact/remove.
2279 D-R- == many chunks to delete / many segments to compact/remove.
2280 Please note that there might be quite some variance in these measure‐
2282 ments. Try multiple measurements and having a otherwise idle machine
2283 (and network, if you use it).
2284
2286 borg help patterns
2287 The path/filenames used as input for the pattern matching start from
2288 the currently active recursion root. You usually give the recursion
2289 root(s) when invoking borg and these can be either relative or absolute
2290 paths.
2291 So, when you give relative/ as root, the paths going into the matcher
2293 will look like relative/.../file.ext. When you give /absolute/ as root,
2294 they will look like /absolute/.../file.ext. This is meant when we talk
2295 about "full path" below.
2296 File patterns support these styles: fnmatch, shell, regular expres‐
2298 sions, path prefixes and path full-matches. By default, fnmatch is used
2299 for --exclude patterns and shell-style is used for the experimental
2300 --pattern option.
2301 If followed by a colon (':') the first two characters of a pattern are
2303 used as a style selector. Explicit style selection is necessary when a
2304 non-default style is desired or when the desired pattern starts with
2305 two alphanumeric characters followed by a colon (i.e. aa:something/*).
2306 Fnmatch, selector fm:
2308 This is the default style for --exclude and --exclude-from.
2309 These patterns use a variant of shell pattern syntax, with '*'
2310 matching any number of characters, '?' matching any single char‐
2311 acter, '[...]' matching any single character specified, includ‐
2312 ing ranges, and '[!...]' matching any character not specified.
2313 For the purpose of these patterns, the path separator (backslash
2314 for Windows and '/' on other systems) is not treated specially.
2315 Wrap meta-characters in brackets for a literal match (i.e. [?]
2316 to match the literal character ?). For a path to match a pat‐
2317 tern, the full path must match, or it must match from the start
2318 of the full path to just before a path separator. Except for the
2319 root path, paths will never end in the path separator when
2320 matching is attempted. Thus, if a given pattern ends in a path
2321 separator, a '*' is appended before matching is attempted.
2322 Shell-style patterns, selector sh:
2324 This is the default style for --pattern and --patterns-from.
2325 Like fnmatch patterns these are similar to shell patterns. The
2326 difference is that the pattern may include **/ for matching zero
2327 or more directory levels, * for matching zero or more arbitrary
2328 characters with the exception of any path separator.
2329 Regular expressions, selector re:
2331 Regular expressions similar to those found in Perl are sup‐
2332 ported. Unlike shell patterns regular expressions are not
2333 required to match the full path and any substring match is suf‐
2334 ficient. It is strongly recommended to anchor patterns to the
2335 start ('^'), to the end ('$') or both. Path separators (back‐
2336 slash for Windows and '/' on other systems) in paths are always
2337 normalized to a forward slash ('/') before applying a pattern.
2338 The regular expression syntax is described in the Python docu‐
2339 mentation for the re module.
2340 Path prefix, selector pp:
2342 This pattern style is useful to match whole sub-directories. The
2343 pattern pp:root/somedir matches root/somedir and everything
2344 therein.
2345 Path full-match, selector pf:
2347 This pattern style is (only) useful to match full paths. This
2348 is kind of a pseudo pattern as it can not have any variable or
2349 unspecified parts - the full path must be given.
2350 pf:root/file.ext matches root/file.txt only.
2351 Implementation note: this is implemented via very time-efficient
2353 O(1) hashtable lookups (this means you can have huge amounts of
2354 such patterns without impacting performance much). Due to that,
2355 this kind of pattern does not respect any context or order. If
2356 you use such a pattern to include a file, it will always be
2357 included (if the directory recursion encounters it). Other
2358 include/exclude patterns that would normally match will be
2359 ignored. Same logic applies for exclude.
2360 NOTE:
2362 re:, sh: and fm: patterns are all implemented on top of the Python
2363 SRE engine. It is very easy to formulate patterns for each of these
2364 types which requires an inordinate amount of time to match paths. If
2365 untrusted users are able to supply patterns, ensure they cannot sup‐
2366 ply re: patterns. Further, ensure that sh: and fm: patterns only
2367 contain a handful of wildcards at most.
2368 Exclusions can be passed via the command line option --exclude. When
2370 used from within a shell the patterns should be quoted to protect them
2371 from expansion.
2372 The --exclude-from option permits loading exclusion patterns from a
2374 text file with one pattern per line. Lines empty or starting with the
2375 number sign ('#') after removing whitespace on both ends are ignored.
2376 The optional style selector prefix is also supported for patterns
2377 loaded from a file. Due to whitespace removal paths with whitespace at
2378 the beginning or end can only be excluded using regular expressions.
2379 To test your exclusion patterns without performing an actual backup you
2381 can run borg create --list --dry-run ....
2382 Examples:
2384 # Exclude '/home/user/file.o' but not '/home/user/file.odt':
2386 $ borg create -e '*.o' backup /
2387 # Exclude '/home/user/junk' and '/home/user/subdir/junk' but
2389 # not '/home/user/importantjunk' or '/etc/junk':
2390 $ borg create -e '/home/*/junk' backup /
2391 # Exclude the contents of '/home/user/cache' but not the directory itself:
2393 $ borg create -e /home/user/cache/ backup /
2394 # The file '/home/user/cache/important' is *not* backed up:
2396 $ borg create -e /home/user/cache/ backup / /home/user/cache/important
2397 # The contents of directories in '/home' are not backed up when their name
2399 # ends in '.tmp'
2400 $ borg create --exclude 're:^/home/[^/]+\.tmp/' backup /
2401 # Load exclusions from file
2403 $ cat >exclude.txt <<EOF
2404 # Comment line
2405 /home/*/junk
2406 *.tmp
2407 fm:aa:something/*
2408 re:^/home/[^/]\.tmp/
2409 sh:/home/*/.thumbnails
2410 EOF
2411 $ borg create --exclude-from exclude.txt backup /
2412 A more general and easier to use way to define filename matching pat‐
2413 terns exists with the experimental --pattern and --patterns-from
2414 options. Using these, you may specify the backup roots (starting
2415 points) and patterns for inclusion/exclusion. A root path starts with
2416 the prefix R, followed by a path (a plain path, not a file pattern). An
2417 include rule starts with the prefix +, an exclude rule starts with the
2418 prefix -, an exclude-norecurse rule starts with !, all followed by a
2419 pattern.
2420 NOTE:
2422 Via --pattern or --patterns-from you can define BOTH inclusion and
2423 exclusion of files using pattern prefixes + and -. With --exclude
2424 and --exlude-from ONLY excludes are defined.
2425 Inclusion patterns are useful to include paths that are contained in an
2427 excluded path. The first matching pattern is used so if an include pat‐
2428 tern matches before an exclude pattern, the file is backed up. If an
2429 exclude-norecurse pattern matches a directory, it won't recurse into it
2430 and won't discover any potential matches for include rules below that
2431 directory.
2432 Note that the default pattern style for --pattern and --patterns-from
2434 is shell style (sh:), so those patterns behave similar to rsync
2435 include/exclude patterns. The pattern style can be set via the P pre‐
2436 fix.
2437 Patterns (--pattern) and excludes (--exclude) from the command line are
2439 considered first (in the order of appearance). Then patterns from
2440 --patterns-from are added. Exclusion patterns from --exclude-from files
2441 are appended last.
2442 Examples:
2444 # backup pics, but not the ones from 2018, except the good ones:
2446 # note: using = is essential to avoid cmdline argument parsing issues.
2447 borg create --pattern=+pics/2018/good --pattern=-pics/2018 repo::arch pics
2448 # use a file with patterns:
2450 borg create --patterns-from patterns.lst repo::arch
2451 The patterns.lst file could look like that:
2453 # "sh:" pattern style is the default, so the following line is not needed:
2455 P sh
2456 R /
2457 # can be rebuild
2458 - /home/*/.cache
2459 # they're downloads for a reason
2460 - /home/*/Downloads
2461 # susan is a nice person
2462 # include susans home
2463 + /home/susan
2464 # don't backup the other home directories
2465 - /home/*
2466 # don't even look in /proc
2467 ! /proc
2468 borg help placeholders
2470 Repository (or Archive) URLs, --prefix, --glob-archives, --comment and
2471 --remote-path values support these placeholders:
2472 {hostname}
2474 The (short) hostname of the machine.
2475 {fqdn} The full name of the machine.
2477 {reverse-fqdn}
2479 The full name of the machine in reverse domain name notation.
2480 {now} The current local date and time, by default in ISO-8601 format.
2482 You can also supply your own format string, e.g.
2483 {now:%Y-%m-%d_%H:%M:%S}
2484 {utcnow}
2486 The current UTC date and time, by default in ISO-8601 format.
2487 You can also supply your own format string, e.g. {utc‐
2488 now:%Y-%m-%d_%H:%M:%S}
2489 {user} The user name (or UID, if no name is available) of the user run‐
2491 ning borg.
2492 {pid} The current process ID.
2494 {borgversion}
2496 The version of borg, e.g.: 1.0.8rc1
2497 {borgmajor}
2499 The version of borg, only the major version, e.g.: 1
2500 {borgminor}
2502 The version of borg, only major and minor version, e.g.: 1.0
2503 {borgpatch}
2505 The version of borg, only major, minor and patch version, e.g.:
2506 1.0.8
2507 If literal curly braces need to be used, double them for escaping:
2509 borg create /path/to/repo::{{literal_text}}
2511 Examples:
2513 borg create /path/to/repo::{hostname}-{user}-{utcnow} ...
2515 borg create /path/to/repo::{hostname}-{now:%Y-%m-%d_%H:%M:%S} ...
2516 borg prune --prefix '{hostname}-' ...
2517 NOTE:
2519 systemd uses a difficult, non-standard syntax for command lines in
2520 unit files (refer to the systemd.unit(5) manual page).
2521 When invoking borg from unit files, pay particular attention to
2523 escaping, especially when using the now/utcnow placeholders, since
2524 systemd performs its own %-based variable replacement even in quoted
2525 text. To avoid interference from systemd, double all percent signs
2526 ({hostname}-{now:%Y-%m-%d_%H:%M:%S} becomes {host‐
2527 name}-{now:%%Y-%%m-%%d_%%H:%%M:%%S}).
2528 borg help compression
2530 It is no problem to mix different compression methods in one repo,
2531 deduplication is done on the source data chunks (not on the compressed
2532 or encrypted data).
2533 If some specific chunk was once compressed and stored into the repo,
2535 creating another backup that also uses this chunk will not change the
2536 stored chunk. So if you use different compression specs for the back‐
2537 ups, whichever stores a chunk first determines its compression. See
2538 also borg recreate.
2539 Compression is lz4 by default. If you want something else, you have to
2541 specify what you want.
2542 Valid compression specifiers are:
2544 none Do not compress.
2546 lz4 Use lz4 compression. Very high speed, very low compression.
2548 (default)
2549 zstd[,L]
2551 Use zstd ("zstandard") compression, a modern wide-range algo‐
2552 rithm. If you do not explicitely give the compression level L
2553 (ranging from 1 to 22), it will use level 3. Archives com‐
2554 pressed with zstd are not compatible with borg < 1.1.4.
2555 zlib[,L]
2557 Use zlib ("gz") compression. Medium speed, medium compression.
2558 If you do not explicitely give the compression level L (ranging
2559 from 0 to 9), it will use level 6. Giving level 0 (means "no
2560 compression", but still has zlib protocol overhead) is usually
2561 pointless, you better use "none" compression.
2562 lzma[,L]
2564 Use lzma ("xz") compression. Low speed, high compression. If
2565 you do not explicitely give the compression level L (ranging
2566 from 0 to 9), it will use level 6. Giving levels above 6 is
2567 pointless and counterproductive because it does not compress
2568 better due to the buffer size used by borg - but it wastes lots
2569 of CPU cycles and RAM.
2570 auto,C[,L]
2572 Use a built-in heuristic to decide per chunk whether to compress
2573 or not. The heuristic tries with lz4 whether the data is com‐
2574 pressible. For incompressible data, it will not use compression
2575 (uses "none"). For compressible data, it uses the given C[,L]
2576 compression - with C[,L] being any valid compression specifier.
2577 Examples:
2579 borg create --compression lz4 REPO::ARCHIVE data
2581 borg create --compression zstd REPO::ARCHIVE data
2582 borg create --compression zstd,10 REPO::ARCHIVE data
2583 borg create --compression zlib REPO::ARCHIVE data
2584 borg create --compression zlib,1 REPO::ARCHIVE data
2585 borg create --compression auto,lzma,6 REPO::ARCHIVE data
2586 borg create --compression auto,lzma ...
2587
2589 There is a borg debug command that has some subcommands which are all
2590 not intended for normal use and potentially very dangerous if used
2591 incorrectly.
2592 For example, borg debug put-obj and borg debug delete-obj will only do
2594 what their name suggests: put objects into repo / delete objects from
2595 repo.
2596 Please note:
2598 · they will not update the chunks cache (chunks index) about the object
2600 · they will not update the manifest (so no automatic chunks index
2602 resync is triggered)
2603 · they will not check whether the object is in use (e.g. before
2605 delete-obj)
2606 · they will not update any metadata which may point to the object
2608 They exist to improve debugging capabilities without direct system
2610 access, e.g. in case you ever run into some severe malfunction. Use
2611 them only if you know what you are doing or if a trusted Borg developer
2612 tells you what to do.
2613 Borg has a --debug-topic TOPIC option to enable specific debugging mes‐
2615 sages. Topics are generally not documented.
2616 A --debug-profile FILE option exists which writes a profile of the main
2618 program's execution to a file. The format of these files is not
2619 directly compatible with the Python profiling tools, since these use
2620 the "marshal" format, which is not intended to be secure (quoting the
2621 Python docs: "Never unmarshal data received from an untrusted or unau‐
2622 thenticated source.").
2623 The borg debug profile-convert command can be used to take a Borg pro‐
2625 file and convert it to a profile file that is compatible with the
2626 Python tools.
2627 Additionally, if the filename specified for --debug-profile ends with
2629 ".pyprof" a Python compatible profile is generated. This is only
2630 intended for local use by developers.
2631
2633 Here are misc. notes about topics that are maybe not covered in enough
2634 detail in the usage section.
2635 --chunker-params
2637 The chunker params influence how input files are cut into pieces
2638 (chunks) which are then considered for deduplication. They also have a
2639 big impact on resource usage (RAM and disk space) as the amount of
2640 resources needed is (also) determined by the total amount of chunks in
2641 the repository (see cache-memory-usage for details).
2642 --chunker-params=10,23,16,4095 results in a fine-grained deduplication|
2644 and creates a big amount of chunks and thus uses a lot of resources to
2645 manage them. This is good for relatively small data volumes and if the
2646 machine has a good amount of free RAM and disk space.
2647 --chunker-params=19,23,21,4095 (default) results in a coarse-grained
2649 deduplication and creates a much smaller amount of chunks and thus uses
2650 less resources. This is good for relatively big data volumes and if the
2651 machine has a relatively low amount of free RAM and disk space.
2652 If you already have made some archives in a repository and you then
2654 change chunker params, this of course impacts deduplication as the
2655 chunks will be cut differently.
2656 In the worst case (all files are big and were touched in between back‐
2658 ups), this will store all content into the repository again.
2659 Usually, it is not that bad though:
2661 · usually most files are not touched, so it will just re-use the old
2663 chunks it already has in the repo
2664 · files smaller than the (both old and new) minimum chunksize result in
2666 only one chunk anyway, so the resulting chunks are same and dedupli‐
2667 cation will apply
2668 If you switch chunker params to save resources for an existing repo
2670 that already has some backup archives, you will see an increasing
2671 effect over time, when more and more files have been touched and stored
2672 again using the bigger chunksize and all references to the smaller
2673 older chunks have been removed (by deleting / pruning archives).
2674 If you want to see an immediate big effect on resource usage, you bet‐
2676 ter start a new repository when changing chunker params.
2677 For more details, see chunker_details.
2679 --noatime / --noctime
2681 You can use these borg create options to not store the respective time‐
2682 stamp into the archive, in case you do not really need it.
2683 Besides saving a little space for the not archived timestamp, it might
2685 also affect metadata stream deduplication: if only this timestamp
2686 changes between backups and is stored into the metadata stream, the
2687 metadata stream chunks won't deduplicate just because of that.
2688 --nobsdflags
2690 You can use this to not query and store (or not extract and set) bsd‐
2691 flags - in case you don't need them or if they are broken somehow for
2692 your fs.
2693 On Linux, dealing with the bsflags needs some additional syscalls.
2695 Especially when dealing with lots of small files, this causes a notica‐
2696 ble overhead, so you can use this option also for speeding up opera‐
2697 tions.
2698 --umask
2700 If you use --umask, make sure that all repository-modifying borg com‐
2701 mands (create, delete, prune) that access the repository in question
2702 use the same --umask value.
2703 If multiple machines access the same repository, this should hold true
2705 for all of them.
2706 --read-special
2708 The --read-special option is special - you do not want to use it for
2709 normal full-filesystem backups, but rather after carefully picking some
2710 targets for it.
2711 The option --read-special triggers special treatment for block and char
2713 device files as well as FIFOs. Instead of storing them as such a device
2714 (or FIFO), they will get opened, their content will be read and in the
2715 backup archive they will show up like a regular file.
2716 Symlinks will also get special treatment if (and only if) they point to
2718 such a special file: instead of storing them as a symlink, the target
2719 special file will get processed as described above.
2720 One intended use case of this is backing up the contents of one or mul‐
2722 tiple block devices, like e.g. LVM snapshots or inactive LVs or disk
2723 partitions.
2724 You need to be careful about what you include when using --read-spe‐
2726 cial, e.g. if you include /dev/zero, your backup will never terminate.
2727 Restoring such files' content is currently only supported one at a time
2729 via --stdout option (and you have to redirect stdout to where ever it
2730 shall go, maybe directly into an existing device file of your choice or
2731 indirectly via dd).
2732 To some extent, mounting a backup archive with the backups of special
2734 files via borg mount and then loop-mounting the image files from inside
2735 the mount point will work. If you plan to access a lot of data in
2736 there, it likely will scale and perform better if you do not work via
2737 the FUSE mount.
2738 Example
2740 Imagine you have made some snapshots of logical volumes (LVs) you want
2741 to backup.
2742 NOTE:
2744 For some scenarios, this is a good method to get "crash-like" con‐
2745 sistency (I call it crash-like because it is the same as you would
2746 get if you just hit the reset button or your machine would abruptly
2747 and completely crash). This is better than no consistency at all
2748 and a good method for some use cases, but likely not good enough if
2749 you have databases running.
2750 Then you create a backup archive of all these snapshots. The backup
2752 process will see a "frozen" state of the logical volumes, while the
2753 processes working in the original volumes continue changing the data
2754 stored there.
2755 You also add the output of lvdisplay to your backup, so you can see the
2757 LV sizes in case you ever need to recreate and restore them.
2758 After the backup has completed, you remove the snapshots again.
2760 $ # create snapshots here
2762 $ lvdisplay > lvdisplay.txt
2763 $ borg create --read-special /path/to/repo::arch lvdisplay.txt /dev/vg0/*-snapshot
2764 $ # remove snapshots here
2765 Now, let's see how to restore some LVs from such a backup.
2767 $ borg extract /path/to/repo::arch lvdisplay.txt
2769 $ # create empty LVs with correct sizes here (look into lvdisplay.txt).
2770 $ # we assume that you created an empty root and home LV and overwrite it now:
2771 $ borg extract --stdout /path/to/repo::arch dev/vg0/root-snapshot > /dev/vg0/root
2772 $ borg extract --stdout /path/to/repo::arch dev/vg0/home-snapshot > /dev/vg0/home
2773 Append-only mode
2775 A repository can be made "append-only", which means that Borg will
2776 never overwrite or delete committed data (append-only refers to the
2777 segment files, but borg will also reject to delete the repository com‐
2778 pletely). This is useful for scenarios where a backup client machine
2779 backups remotely to a backup server using borg serve, since a hacked
2780 client machine cannot delete backups on the server permanently.
2781 To activate append-only mode, set append_only to 1 in the repository
2783 config:
2784 borg config /path/to/repo append_only 1
2786 Note that you can go back-and-forth between normal and append-only
2788 operation with borg config; it's not a "one way trip."
2789 In append-only mode Borg will create a transaction log in the transac‐
2791 tions file, where each line is a transaction and a UTC timestamp.
2792 In addition, borg serve can act as if a repository is in append-only
2794 mode with its option --append-only. This can be very useful for
2795 fine-tuning access control in .ssh/authorized_keys:
2796 command="borg serve --append-only ..." ssh-rsa <key used for not-always-trustable backup clients>
2798 command="borg serve ..." ssh-rsa <key used for backup management>
2799 Running borg init via a borg serve --append-only server will not create
2801 an append-only repository. Running borg init --append-only creates an
2802 append-only repository regardless of server settings.
2803 Example
2805 Suppose an attacker remotely deleted all backups, but your repository
2806 was in append-only mode. A transaction log in this situation might look
2807 like this:
2808 transaction 1, UTC time 2016-03-31T15:53:27.383532
2810 transaction 5, UTC time 2016-03-31T15:53:52.588922
2811 transaction 11, UTC time 2016-03-31T15:54:23.887256
2812 transaction 12, UTC time 2016-03-31T15:55:54.022540
2813 transaction 13, UTC time 2016-03-31T15:55:55.472564
2814 From your security logs you conclude the attacker gained access at
2816 15:54:00 and all the backups where deleted or replaced by compromised
2817 backups. From the log you know that transactions 11 and later are com‐
2818 promised. Note that the transaction ID is the name of the last file in
2819 the transaction. For example, transaction 11 spans files 6 to 11.
2820 In a real attack you'll likely want to keep the compromised repository
2822 intact to analyze what the attacker tried to achieve. It's also a good
2823 idea to make this copy just in case something goes wrong during the
2824 recovery. Since recovery is done by deleting some files, a hard link
2825 copy (cp -al) is sufficient.
2826 The first step to reset the repository to transaction 5, the last
2828 uncompromised transaction, is to remove the hints.N, index.N and
2829 ``
2830 integrity.N``files in the repository (these files are always expend‐
2831 able). In this example N is 13.
2832 Then remove or move all segment files from the segment directories in
2834 data/ starting with file 6:
2835 rm data/**/{6..13}
2837 That's all to do in the repository.
2839 If you want to access this rollbacked repository from a client that
2841 already has a cache for this repository, the cache will reflect a newer
2842 repository state than what you actually have in the repository now,
2843 after the rollback.
2844 Thus, you need to clear the cache:
2846 borg delete --cache-only repo
2848 The cache will get rebuilt automatically. Depending on repo size and
2850 archive count, it may take a while.
2851 You also will need to remove ~/.config/borg/security/REPOID/mani‐
2853 fest-timestamp.
2854 Drawbacks
2856 As data is only appended, and nothing removed, commands like prune or
2857 delete won't free disk space, they merely tag data as deleted in a new
2858 transaction.
2859 Be aware that as soon as you write to the repo in non-append-only mode
2861 (e.g. prune, delete or create archives from an admin machine), it will
2862 remove the deleted objects permanently (including the ones that were
2863 already marked as deleted, but not removed, in append-only mode). Auto‐
2864 mated edits to the repository (such as a cron job running borg prune)
2865 will render append-only mode moot if data is deleted.
2866 Even if an archive appears to be available, it is possible an attacker
2868 could delete just a few chunks from an archive and silently corrupt its
2869 data. While in append-only mode, this is reversible, but borg check
2870 should be run before a writing/pruning operation on an append-only
2871 repository to catch accidental or malicious corruption:
2872 # run without append-only mode
2874 borg check --verify-data repo
2875 Aside from checking repository & archive integrity you may want to also
2877 manually check backups to ensure their content seems correct.
2878 Further considerations
2880 Append-only mode is not respected by tools other than Borg. rm still
2881 works on the repository. Make sure that backup client machines only get
2882 to access the repository via borg serve.
2883 Ensure that no remote access is possible if the repository is temporar‐
2885 ily set to normal mode for e.g. regular pruning.
2886 Further protections can be implemented, but are outside of Borg's
2888 scope. For example, file system snapshots or wrapping borg serve to set
2889 special permissions or ACLs on new data files.
2890 SSH batch mode
2892 When running Borg using an automated script, ssh might still ask for a
2893 password, even if there is an SSH key for the target server. Use this
2894 to make scripts more robust:
2895 export BORG_RSH='ssh -oBatchMode=yes'
2897
2899 The Borg Collective (see AUTHORS file)
2900
2902 2010-2020 Jonas Borgström, 2015-2019 The Borg Collective (see AUTHORS
2903 file)
29041.1.11 2020-03-08 BORG(1)