1 bup: It backs things up
2 =======================
4 bup is a program that backs things up. It's short for "backup." Can you
5 believe that nobody else has named an open source program "bup" after all
8 Despite its unassuming name, bup is pretty cool. To give you an idea of
9 just how cool it is, I wrote you this poem:
12 What rhymes with awesome?
14 But that's irrelevant.
16 Hmm. Did that help? Maybe prose is more useful after all.
19 Reasons bup is awesome
20 ----------------------
22 bup has a few advantages over other backup software:
24 - It uses a rolling checksum algorithm (similar to rsync) to split large
25 files into chunks. The most useful result of this is you can backup huge
26 virtual machine (VM) disk images, databases, and XML files incrementally,
27 even though they're typically all in one huge file, and not use tons of
28 disk space for multiple versions.
30 - It uses the packfile format from git (the open source version control
31 system), so you can access the stored data even if you don't like bup's
34 - Unlike git, it writes packfiles *directly* (instead of having a separate
35 garbage collection / repacking stage) so it's fast even with gratuitously
36 huge amounts of data. bup's improved index formats also allow you to
37 track far more filenames than git (millions) and keep track of far more
38 objects (hundreds or thousands of gigabytes).
40 - Data is "automagically" shared between incremental backups without having
41 to know which backup is based on which other one - even if the backups
42 are made from two different computers that don't even know about each
43 other. You just tell bup to back stuff up, and it saves only the minimum
44 amount of data needed.
46 - You can back up directly to a remote bup server, without needing tons of
47 temporary disk space on the computer being backed up. And if your backup
48 is interrupted halfway through, the next run will pick up where you left
49 off. And it's easy to set up a bup server: just install bup on any
50 machine where you have ssh access.
52 - Bup can use "par2" redundancy to recover corrupted backups even if your
53 disk has undetected bad sectors.
55 - Even when a backup is incremental, you don't have to worry about
56 restoring the full backup, then each of the incrementals in turn; an
57 incremental backup *acts* as if it's a full backup, it just takes less
60 - You can mount your bup repository as a FUSE filesystem and access the
61 content that way, and even export it over Samba.
63 - It's written in python (with some C parts to make it faster) so it's easy
64 for you to extend and maintain.
67 Reasons you might want to avoid bup
68 -----------------------------------
70 - It's not remotely as well tested as something like tar, so it's
71 more likely to eat your data. It's also missing some
72 probably-critical features, though fewer than it used to be.
74 - It requires python >= 2.6, a C compiler, and an installed git
75 version >= 1.5.6. It also requires par2 if you want fsck to be
76 able to generate the information needed to recover from some types
79 - It currently only works on Linux, FreeBSD, NetBSD, OS X >= 10.4,
80 Solaris, or Windows (with Cygwin, and maybe with WSL). Patches to
81 support other platforms are welcome.
83 - Any items in "Things that are stupid" below.
86 Notable changes introduced by a release
87 =======================================
89 - <a href="note/0.29.1-from-0.29.md">Changes in 0.29.1 as compared to 0.29</a>
90 - <a href="note/0.29-from-0.28.1.md">Changes in 0.29 as compared to 0.28.1</a>
91 - <a href="note/0.28.1-from-0.28.md">Changes in 0.28.1 as compared to 0.28</a>
92 - <a href="note/0.28-from-0.27.1.md">Changes in 0.28 as compared to 0.27.1</a>
93 - <a href="note/0.27.1-from-0.27.md">Changes in 0.27.1 as compared to 0.27</a>
102 - Check out the bup source code using git:
104 git clone https://github.com/bup/bup
106 - This will leave you on the master branch, which is perfect if you
107 would like to help with development, but if you'd just like to use
108 bup, please check out the latest stable release like this:
112 You can see the latest stable release here:
113 https://github.com/bup/bup/releases.
115 - Install the required python libraries (including the development
118 On very recent Debian/Ubuntu versions, this may be sufficient (run
121 apt-get build-dep bup
123 Otherwise try this (substitute python2.6-dev if you have an older
126 apt-get install python2.7-dev python-fuse
127 apt-get install python-pyxattr python-pylibacl
128 apt-get install linux-libc-dev
129 apt-get install acl attr
130 apt-get install python-tornado # optional
132 On CentOS (for CentOS 6, at least), this should be sufficient (run
135 yum groupinstall "Development Tools"
136 yum install python python-devel
137 yum install fuse-python pyxattr pylibacl
138 yum install perl-Time-HiRes
140 In addition to the default CentOS repositories, you may need to add
141 RPMForge (for fuse-python) and EPEL (for pyxattr and pylibacl).
143 On Cygwin, install python, make, rsync, and gcc4.
145 If you would like to use the optional bup web server on systems
146 without a tornado package, you may want to try this:
150 - Build the python module and symlinks:
158 The tests should pass. If they don't pass for you, stop here and
159 send an email to bup-list@googlegroups.com. Though if there are
160 symbolic links along the current working directory path, the tests
161 may fail. Running something like this before "make test" should
162 sidestep the problem:
166 - You can install bup via "make install", and override the default
167 destination with DESTDIR and PREFIX.
169 Files are normally installed to "$DESTDIR/$PREFIX" where DESTDIR is
170 empty by default, and PREFIX is set to /usr/local. So if you wanted to
171 install bup to /opt/bup, you might do something like this:
173 make install DESTDIR=/opt/bup PREFIX=''
175 - The Python executable that bup will use is chosen by ./configure,
176 which will search for a reasonable version unless PYTHON is set in
177 the environment, in which case, bup will use that path. You can
178 see which Python executable was chosen by looking at the
179 configure output, or examining cmd/python-cmd.sh, and you can
180 change the selection by re-running ./configure.
185 Binary packages of bup are known to be built for the following OSes:
188 http://packages.debian.org/search?searchon=names&keywords=bup
190 http://packages.ubuntu.com/search?searchon=names&keywords=bup
191 - pkgsrc (NetBSD, Dragonfly, and others)
192 http://pkgsrc.se/sysutils/bup
193 http://cvsweb.netbsd.org/bsdweb.cgi/pkgsrc/sysutils/bup/
195 https://www.archlinux.org/packages/?sort=&q=bup
197 https://apps.fedoraproject.org/packages/bup
203 - Get help for any bup command:
212 - Initialize the default BUP_DIR (~/.bup -- you can choose another by
213 either specifying `bup -d DIR ...` or setting the `BUP_DIR`
214 environment variable for a command):
218 - Make a local backup (-v or -vv will increase the verbosity):
221 bup save -n local-etc /etc
223 - Restore a local backup to ./dest:
225 bup restore -C ./dest local-etc/latest/etc
228 - Look at how much disk space your backup took:
232 - Make another backup (which should be mostly identical to the last one;
233 notice that you don't have to *specify* that this backup is incremental,
234 it just saves space automatically):
237 bup save -n local-etc /etc
239 - Look how little extra space your second backup used (on top of the first):
243 - Get a list of your previous backups:
247 - Restore your first backup again:
249 bup restore -C ./dest-2 local-etc/2013-11-23-11195/etc
251 - Make a backup to a remote server which must already have the 'bup' command
252 somewhere in its PATH (see /etc/profile, etc/environment, ~/.profile, or
253 ~/.bashrc), and be accessible via ssh.
254 Make sure to replace SERVERNAME with the actual hostname of your server:
256 bup init -r SERVERNAME:path/to/remote-bup-dir
258 bup save -r SERVERNAME:path/to/remote-bup-dir -n local-etc /etc
260 - Make a remote backup to ~/.bup on SERVER:
263 bup save -r SERVER: -n local-etc /etc
265 - See what saves are available in ~/.bup on SERVER:
269 - Restore the remote backup to ./dest:
271 bup restore -r SERVER: -C ./dest local-etc/latest/etc
274 - Defend your backups from death rays (OK fine, more likely from the
275 occasional bad disk block). This writes parity information
276 (currently via par2) for all of the existing data so that bup may
277 be able to recover from some amount of repository corruption:
281 - Use split/join instead of index/save/restore. Try making a local
284 tar -cvf - /etc | bup split -n local-etc -vv
286 - Try restoring the tarball:
288 bup join local-etc | tar -tf -
290 - Look at how much disk space your backup took:
294 - Make another tar backup:
296 tar -cvf - /etc | bup split -n local-etc -vv
298 - Look at how little extra space your second backup used on top of
303 - Restore the first tar backup again (the ~1 is git notation for "one
304 older than the most recent"):
306 bup join local-etc~1 | tar -tf -
308 - Get a list of your previous split-based backups:
310 GIT_DIR=~/.bup git log local-etc
312 - Save a tar archive to a remote server (without tar -z to facilitate
315 tar -cvf - /etc | bup split -r SERVERNAME: -n local-etc -vv
317 - Restore the archive:
319 bup join -r SERVERNAME: local-etc | tar -tf -
321 That's all there is to it!
327 - FreeBSD's default 'make' command doesn't like bup's Makefile. In order to
328 compile the code, run tests and install bup, you need to install GNU Make
329 from the port named 'gmake' and use its executable instead in the commands
330 seen above. (i.e. 'gmake test' runs bup's test suite)
332 - Python's development headers are automatically installed with the 'python'
333 port so there's no need to install them separately.
335 - To use the 'bup fuse' command, you need to install the fuse kernel module
336 from the 'fusefs-kmod' port in the 'sysutils' section and the libraries from
337 the port named 'py-fusefs' in the 'devel' section.
339 - The 'par2' command can be found in the port named 'par2cmdline'.
341 - In order to compile the documentation, you need pandoc which can be found in
342 the port named 'hs-pandoc' in the 'textproc' section.
345 Notes on NetBSD/pkgsrc
346 ----------------------
348 - See pkgsrc/sysutils/bup, which should be the most recent stable
349 release and includes man pages. It also has a reasonable set of
350 dependencies (git, par2, py-fuse-bindings).
352 - The "fuse-python" package referred to is hard to locate, and is a
353 separate tarball for the python language binding distributed by the
354 fuse project on sourceforge. It is available as
355 pkgsrc/filesystems/py-fuse-bindings and on NetBSD 5, "bup fuse"
358 - "bup fuse" presents every directory/file as inode 0. The directory
359 traversal code ("fts") in NetBSD's libc will interpret this as a
360 cycle and error out, so "ls -R" and "find" will not work.
362 - There is no support for ACLs. If/when some entrprising person
363 fixes this, adjust t/compare-trees.
369 - There is no support for ACLs. If/when some enterprising person
370 fixes this, adjust t/compare-trees.
372 - In t/test.sh, two tests have been disabled. These tests check to
373 see that repeated saves produce identical trees and that an
374 intervening index doesn't change the SHA1. Apparently Cygwin has
375 some unusual behaviors with respect to access times (that probably
376 warrant further investigation). Possibly related:
377 http://cygwin.com/ml/cygwin/2007-06/msg00436.html
383 - There is no support for ACLs. If/when some enterprising person
384 fixes this, adjust t/compare-trees.
393 bup stores its data in a git-formatted repository. Unfortunately, git
394 itself doesn't actually behave very well for bup's use case (huge numbers of
395 files, files with huge sizes, retaining file permissions/ownership are
396 important), so we mostly don't use git's *code* except for a few helper
397 programs. For example, bup has its own git packfile writer written in
400 Basically, 'bup split' reads the data on stdin (or from files specified on
401 the command line), breaks it into chunks using a rolling checksum (similar to
402 rsync), and saves those chunks into a new git packfile. There is at least one
403 git packfile per backup.
405 When deciding whether to write a particular chunk into the new packfile, bup
406 first checks all the other packfiles that exist to see if they already have that
407 chunk. If they do, the chunk is skipped.
409 git packs come in two parts: the pack itself (*.pack) and the index (*.idx).
410 The index is pretty small, and contains a list of all the objects in the
411 pack. Thus, when generating a remote backup, we don't have to have a copy
412 of the packfiles from the remote server: the local end just downloads a copy
413 of the server's *index* files, and compares objects against those when
414 generating the new pack, which it sends directly to the server.
416 The "-n" option to 'bup split' and 'bup save' is the name of the backup you
417 want to create, but it's actually implemented as a git branch. So you can
418 do cute things like checkout a particular branch using git, and receive a
419 bunch of chunk files corresponding to the file you split.
421 If you use '-b' or '-t' or '-c' instead of '-n', bup split will output a
422 list of blobs, a tree containing that list of blobs, or a commit containing
423 that tree, respectively, to stdout. You can use this to construct your own
424 scripts that do something with those values.
429 'bup index' walks through your filesystem and updates a file (whose name is,
430 by default, ~/.bup/bupindex) to contain the name, attributes, and an
431 optional git SHA1 (blob id) of each file and directory.
433 'bup save' basically just runs the equivalent of 'bup split' a whole bunch
434 of times, once per file in the index, and assembles a git tree
435 that contains all the resulting objects. Among other things, that makes
436 'git diff' much more useful (compared to splitting a tarball, which is
437 essentially a big binary blob). However, since bup splits large files into
438 smaller chunks, the resulting tree structure doesn't *exactly* correspond to
439 what git itself would have stored. Also, the tree format used by 'bup save'
440 will probably change in the future to support storing file ownership, more
441 complex file permissions, and so on.
443 If a file has previously been written by 'bup save', then its git blob/tree
444 id is stored in the index. This lets 'bup save' avoid reading that file to
445 produce future incremental backups, which means it can go *very* fast unless
446 a lot of files have changed.
449 Things that are stupid for now but which we'll fix later
450 ========================================================
452 Help with any of these problems, or others, is very welcome. Join the
453 mailing list (see below) if you'd like to help.
455 - 'bup save' and 'bup restore' have immature metadata support.
457 On the plus side, they actually do have support now, but it's new,
458 and not remotely as well tested as tar/rsync/whatever's. However,
459 you have to start somewhere, and as of 0.25, we think it's ready
460 for more general use. Please let us know if you have any trouble.
462 Also, if any strip or graft-style options are specified to 'bup
463 save', then no metadata will be written for the root directory.
464 That's obviously less than ideal.
466 - bup is overly optimistic about mmap. Right now bup just assumes
467 that it can mmap as large a block as it likes, and that mmap will
468 never fail. Yeah, right... If nothing else, this has failed on
469 32-bit architectures (and 31-bit is even worse -- looking at you,
472 To fix this, we might just implement a FakeMmap[1] class that uses
473 normal file IO and handles all of the mmap methods[2] that bup
474 actually calls. Then we'd swap in one of those whenever mmap
477 This would also require implementing some of the methods needed to
478 support "[]" array access, probably at a minimum __getitem__,
479 __setitem__, and __setslice__ [3].
481 [1] http://comments.gmane.org/gmane.comp.sysutils.backup.bup/613
482 [2] http://docs.python.org/2/library/mmap.html
483 [3] http://docs.python.org/2/reference/datamodel.html#emulating-container-types
485 - 'bup index' is slower than it should be.
487 It's still rather fast: it can iterate through all the filenames on my
488 600,000 file filesystem in a few seconds. But it still needs to rewrite
489 the entire index file just to add a single filename, which is pretty
490 nasty; it should just leave the new files in a second "extra index" file
493 - bup could use inotify for *really* efficient incremental backups.
495 You could even have your system doing "continuous" backups: whenever a
496 file changes, we immediately send an image of it to the server. We could
497 give the continuous-backup process a really low CPU and I/O priority so
498 you wouldn't even know it was running.
500 - bup only has experimental support for pruning old backups.
502 While you should now be able to drop old saves and branches with
503 `bup rm`, and reclaim the space occupied by data that's no longer
504 needed by other backups with `bup gc`, these commands are
505 experimental, and should be handled with great care. See the
506 man pages for more information.
508 Unless you want to help test the new commands, one possible
509 workaround is to just start a new BUP_DIR occasionally,
510 i.e. bup-2013, bup-2014...
512 - bup has never been tested on anything but Linux, FreeBSD, NetBSD,
513 OS X, and Windows+Cygwin.
515 There's nothing that makes it *inherently* non-portable, though, so
516 that's mostly a matter of someone putting in some effort. (For a
517 "native" Windows port, the most annoying thing is the absence of ssh in
518 a default Windows installation.)
520 - bup needs better documentation.
522 According to an article about bup in Linux Weekly News
523 (https://lwn.net/Articles/380983/), "it's a bit short on examples and
524 a user guide would be nice." Documentation is the sort of thing that
525 will never be great unless someone from outside contributes it (since
526 the developers can never remember which parts are hard to understand).
528 - bup is "relatively speedy" and has "pretty good" compression.
530 ...according to the same LWN article. Clearly neither of those is good
531 enough. We should have awe-inspiring speed and crazy-good compression.
532 Must work on that. Writing more parts in C might help with the speed.
536 Actually, that's not stupid, but you might consider it a
537 limitation. See the ["Related Projects"](https://bup.github.io/)
538 list for some possible options.
543 bup has an extensive set of man pages. Try using 'bup help' to get
544 started, or use 'bup help SUBCOMMAND' for any bup subcommand (like split,
545 join, index, save, etc.) to get details on that command.
547 For further technical details, please see ./DESIGN.
553 bup is a work in progress and there are many ways it can still be improved.
554 If you'd like to contribute patches, ideas, or bug reports, please join the
557 You can find the mailing list archives here:
559 http://groups.google.com/group/bup-list
561 and you can subscribe by sending a message to:
563 bup-list+subscribe@googlegroups.com
565 Please see <a href="HACKING">./HACKING</a> for
566 additional information, i.e. how to submit patches (hint - no pull
567 requests), how we handle branches, etc.