1 mailto(rsync-bugs@samba.anu.edu.au)
2 manpage(rsync)(1)(13 May 1998)()()
3 manpagename(rsync)(faster, flexible replacement for rcp)
6 rsync [options] [user@]host:path path
8 rsync [options] path [user@]host:path
10 rsync [options] path path
12 rsync [options] [user@]host::module[/path] path
14 rsync [options] path [user@]host::module[/path]
16 rsync [options] rsync://host/module/path path
20 rsync is a program that behaves in much the same way that rcp does,
21 but has many more options and uses the rsync remote-update protocol to
22 greatly speedup file transfers when the destination file already
25 The rsync remote-update protocol allows rsync to transfer just the
26 differences between two sets of files across the network link, using
27 an efficient checksum-search algorithm described in the technical
28 report that accompanies this package.
30 Some of the additional features of rsync are:
33 it() support for copying links, devices, owners, groups and permissions
34 it() exclude and exclude-from options similar to GNU tar
35 it() a CVS exclude mode for ignoring the same files that CVS would ignore
36 it() can use any transparent remote shell, including rsh or ssh
37 it() does not require root privileges
38 it() pipelining of file transfers to minimize latency costs
39 it() support for anonymous or authenticated rsync servers (ideal for
43 manpagesection(GENERAL)
45 There are six different ways of using rsync. They are:
48 it() for copying local files. This is invoked when neither
49 source nor destination path contains a : separator
51 it() for copying from the local machine to a remote machine using
52 a remote shell program as the transport (such as rsh or
53 ssh). This is invoked when the destination path contains a
56 it() for copying from a remote machine to the local machine
57 using a remote shell program. This is invoked when the local path
58 contains a : separator.
60 it() for copying from a remote rsync server to the local
61 machine. This is invoked when the source path contains a ::
62 separator. You can also use a rsync:// URL if no username
65 it() for copying from the local machine to a remote rsync
66 server. This is invoked when the destination path contains a ::
69 it() for listing files on a remote machine. This is done the
70 same way as rsync transfers except that you leave off the
74 Note that in all cases at least one of the source and destination
79 See the file README for installation instructions.
81 Once installed you can use rsync to any machine that you can use rsh
82 to. rsync uses rsh for its communications, unless both the source and
83 destination are local.
85 You can also specify a alternative to rsh, by either using the -e
86 command line option, or by setting the RSYNC_RSH environment variable.
88 One common substitute is to use ssh, which offers a high degree of
91 Note that rsync must be installed on both the source and destination
96 You use rsync in the same way you use rcp. You must specify a source
97 and a destination, one of which may be remote.
99 Perhaps the best way to explain the syntax is some examples:
101 quote(rsync *.c foo:src/)
103 this would transfer all files matching the pattern *.c from the
104 current directory to the directory src on the machine foo. If any of
105 the files already exist on the remote system then the rsync
106 remote-update protocol is used to update the file by sending only the
107 differences. See the tech report for details.
109 quote(rsync -avz foo:src/bar /data/tmp)
111 recursively transfer all files from the directory src/bar on the
112 machine foo into the /data/tmp/bar directory on the local machine. The
113 files are transferred in "archive" mode, which ensures that symbolic
114 links, devices, attributes, permissions, ownerships etc are preserved
115 in the transfer. Additionally compression will be used to reduce the
116 size of data portions of the transfer.
118 quote(rsync -avz foo:src/bar/ /data/tmp)
120 With a trailing slash on the source this behavior changes to transfer
121 all files from the directory src/bar on the machine foo into the
122 /data/tmp/. With a trailing / on a source name it means "copy the
123 contents of this directory". Without a trailing slash it means "copy
124 the directory". This difference becomes particularly important when
125 using the --delete option.
127 You can also use rsync in local-only mode, where both the source and
128 destination don't have a ':' in the name. In this case it behaves like
129 an improved copy command.
132 manpagesection(CONNECTING TO AN RSYNC SERVER)
134 It is also possible to use rsync without using rsh or ssh as the
135 transport. In this case you will connect to a remote rsync server
136 running on TCP port 873.
138 Using rsync in this was is the same as using it with rsh or ssh except
142 it() you use a double colon :: instead of a single colon to
143 separate the hostname from the path.
145 it() the remote server may print a message of the day when you
148 it() if you specify no path name on the remote server then the
149 list of accessible paths on the server will be shown.
151 it() if you specify no local destination then a listing of the
152 specified files on the remote server is provided
155 Some paths on the remote server may require authentication. If so then
156 you will receive a password prompt when you connect. You can avoid the
157 password prompt by setting the environment variable RSYNC_PASSWORD to
158 the password you want to use. This may be useful when scripting rsync.
160 manpagesection(RUNNING AN RSYNC SERVER)
162 An rsync server is configured using a config file which by default is
163 called /etc/rsyncd.conf. Please see the rsyncd.conf(5) man page for more
166 manpagesection(EXAMPLES)
168 Here are some examples of how I use rsync.
170 To backup my wife's home directory, which consists of large MS word
171 files and mail folders I use a cron job that runs
173 quote(rsync -Cavz . arvidsjaur:backup)
175 each night over a PPP link to a duplicate directory on my machine
178 To synchronize my samba source trees I use the following Makefile
182 rsync -avuzb --exclude '*~' samba:samba/ .
185 rsync -Cavuzb . samba:samba/
189 this allows me to sync with a CVS directory at the other end of the
190 link. I then do cvs operations on the remote machine, which saves a
191 lot of time as the remote cvs protocol isn't very efficient.
193 I mirror a directory between my "old" and "new" ftp sites with the
196 quote(rsync -az -e ssh --delete ~ftp/pub/samba/ nimbus:"~ftp/pub/tridge/samba")
198 this is launched from cron every few hours.
200 manpagesection(OPTIONS SUMMARY)
202 Here is a short summary of the options avalable in rsync. Please refer
203 to the detailed description below for a complete description.
206 Usage: rsync [OPTION]... SRC [USER@]HOST:DEST
207 or rsync [OPTION]... [USER@]HOST:SRC DEST
208 or rsync [OPTION]... SRC DEST
209 or rsync [OPTION]... [USER@]HOST::SRC DEST
210 or rsync [OPTION]... SRC [USER@]HOST::DEST
213 -v, --verbose increase verbosity
214 -c, --checksum always checksum
215 -a, --archive archive mode
216 -r, --recursive recurse into directories
217 -R, --relative use relative path names
218 -b, --backup make backups (default ~ extension)
219 -u, --update update only (don't overwrite newer files)
220 -l, --links preserve soft links
221 -L, --copy-links treat soft links like regular files
222 --safe-links ignore links outside the destination tree
223 -H, --hard-links preserve hard links
224 -p, --perms preserve permissions
225 -o, --owner preserve owner (root only)
226 -g, --group preserve group
227 -D, --devices preserve devices (root only)
228 -t, --times preserve times
229 -S, --sparse handle sparse files efficiently
230 -n, --dry-run show what would have been transferred
231 -W, --whole-file copy whole files, no incremental checks
232 -x, --one-file-system don't cross filesystem boundaries
233 -B, --block-size=SIZE checksum blocking size
234 -e, --rsh=COMMAND specify rsh replacement
235 --rsync-path=PATH specify path to rsync on the remote machine
236 -C, --cvs-exclude auto ignore files in the same way CVS does
237 --delete delete files that don't exist on the sending side
238 --partial keep partially transferred files
239 --force force deletion of directories even if not empty
240 --numeric-ids don't map uid/gid values by user/group name
241 --timeout=TIME set IO timeout in seconds
242 -I, --ignore-times don't exclude files that match length and time
243 -T --temp-dir=DIR create temporary files in directory DIR
244 --compare-dest=DIR also compare destination files relative to DIR
245 -z, --compress compress file data
246 --exclude=PATTERN exclude file FILE
247 --exclude-from=PATTERN exclude files listed in FILE
248 --include=PATTERN don't exclude file FILE
249 --include-from=PATTERN don't exclude files listed in FILE
250 --suffix=SUFFIX override backup suffix
251 --version print version number
252 --daemon run as a rsync daemon
253 --config=FILE specify alternate rsyncd.conf file
254 --port=PORT specify alternate rsyncd port number
255 --stats give some file transfer stats
256 --progress show progress during transfer
257 -h, --help show this help screen
262 rsync uses the GNU long options package. Many of the command line
263 options have two variants, one short and one long. These are shown
264 below separated by commas. Some options only have a long variant.
267 dit(bf(-h, --help)) Print a short help page describing the options
270 dit(bf(--version)) print the rsync version number and exit
272 dit(bf(-v, --verbose)) This option increases the amount of information you
273 are given during the transfer. By default rsync works silently. A
274 single -v will give you information about what files are being
275 transferred and a brief summary at the end. Two -v flags will give you
276 information on what files are being skipped and slightly more
277 information at the end. More than two -v flags should only be used if
278 you are debugging rsync
280 dit(bf(-I, --ignore-times)) Normally rsync will skip any files that are
281 already the same length and have the same time-stamp. This option turns
284 dit(bf(-c, --checksum)) This forces the sender to checksum all files using
285 a 128-bit MD4 checksum before transfer. The checksum is then
286 explicitly checked on the receiver and any files of the same name
287 which already exist and have the same checksum and size on the
288 receiver are skipped. This option can be quite slow.
290 dit(bf(-a, --archive)) This is equivalent to -rlptDg. It is a quick way
291 of saying I want recursion and want to preserve everything.
293 Note: if the user launching rsync is root then the -o option (preserve
294 uid) is also implied.
296 dit(bf(-r, --recursive)) This tells rsync to copy directories recursively
298 dit(bf(-R, --relative)) Use relative paths. This means that the full path
299 names specified on the command line are sent to the server rather than
300 just the last parts of the filenames. This is particularly useful when
301 you want to sent several different directories at the same time. For
302 example if you used the command
304 verb(rsync foo/bar/foo.c remote:/tmp/)
306 then this would create a file called foo.c in /tmp/ on the remote
307 machine. If instead you used
309 verb(rsync -R foo/bar/foo.c remote:/tmp/)
311 then a file called /tmp/foo/bar/foo.c would be created on the remote
312 machine. The full path name is preserved.
314 dit(bf(-b, --backup)) With this option preexisting destination files are
315 renamed with a ~ extension as each file is transferred. You can
316 control the backup suffix using the --suffix option.
318 dit(bf(-u, --update)) This forces rsync to skip any files for which the
319 destination file already exists and has a date later than the source
322 dit(bf(-l, --links)) This tells rsync to recreate symbolic links on the
323 remote system to be the same as the local system. Without this
324 option all symbolic links are skipped.
326 dit(bf(-L, --copy-links)) This tells rsync to treat symbolic links just
329 dit(bf(--safe-links)) This tells rsync to ignore any symbolic links
330 which point outside the destination tree. All absolute symlinks are
331 also ignored. Using this option in conjunction with --relative may
332 give unexpecetd results.
334 dit(bf(-H, --hard-links)) This tells rsync to recreate hard links on
335 the remote system to be the same as the local system. Without this
336 option hard links are treated like regular files.
338 Note that rsync can only detect hard links if both parts of the link
339 are in the list of files being sent.
341 This option can be quite slow, so only use it if you need it.
343 dit(bf(-W, --whole-file)) With this option the incremental rsync algorithm
344 is not used and the whole file is sent as-is instead. This may be
345 useful when using rsync with a local machine.
347 dit(bf(--partial)) By default rsync will delete any partially
348 transferred file if the transfer is interrupted. In some circumstances
349 it is more desirable to keep partially transferred files. Using the
350 --partial option tells rsync to keep the partial file which should
351 make a subsequent transfer of the rest of the file much faster.
353 dit(bf(-p, --perms)) This option causes rsync to update the remote
354 permissions to be the same as the local permissions.
356 dit(bf(-o, --owner)) This option causes rsync to update the remote owner
357 of the file to be the same as the local owner. This is only available
358 to the super-user. Note that if the source system is a daemon using chroot,
359 the --numeric-ids option is implied because the source system cannot get
360 access to the user names.
362 dit(bf(-g, --group)) This option causes rsync to update the remote group
363 of the file to be the same as the local group. Note that if the source
364 system is a daemon using chroot, the --numeric-ids option is implied because
365 the source system cannot get access to the group names.
367 dit(bf(-D, --devices)) This option causes rsync to transfer character and
368 block device information to the remote system to recreate these
369 devices. This option is only available to the super-user.
371 dit(bf(-t, --times)) This tells rsync to transfer modification times along
372 with the files and update them on the remote system
374 dit(bf(-n, --dry-run)) This tells rsync to not do any file transfers,
375 instead it will just report the actions it would have taken.
377 dit(bf(-S, --sparse)) Try to handle sparse files efficiently so they take
378 up less space on the destination.
380 NOTE: Don't use this option when the destination is a Solaris "tmpfs"
381 filesystem. It doesn't seem to handle seeks over null regions
382 correctly and ends up corrupting the files.
384 dit(bf(-x, --one-file-system)) This tells rsync not to cross filesystem
385 boundaries when recursing. This is useful for transferring the
386 contents of only one filesystem.
388 dit(bf(--delete)) This tells rsync to delete any files on the receiving
389 side that aren't on the sending side. This option can be dangerous if
392 It is a very good idea to run first using the dry run option (-n) to
393 see what files would be deleted to make sure important files aren't
396 rsync 1.6.4 changed the behavior of --delete to make it less
397 dangerous. rsync now only scans directories on the receiving side
398 that are explicitly transferred from the sending side. Only files in
399 these directories are deleted.
401 Still, it is probably easy to get burnt with this option. The moral
402 of the story is to use the -n option until you get used to the
403 behavior of --delete.
405 NOTE: It also may delete files on the destination if the sending side
406 can't open them or stat them. This is a bug that hopefully will be
407 fixed in a future release.
409 dit(bf(--force)) This options tells rsync to delete directories even if
410 they are not empty. This applies to both the --delete option and to
411 cases where rsync tries to copy a normal file but the destination
412 contains a directory of the same name. Normally rsync will refuse to
413 do a recursive directory deletion in such cases, by using --force
414 the recursive deletion will be done.
416 Use this option with caution!
418 dit(bf(-B , --block_size BLOCKSIZE)) This controls the block size used in
419 the rsync algorithm. See the technical report for details.
421 dit(bf(-e, --rsh COMMAND)) This option allows you to choose an alternative
422 remote shell program to use for communication between the local and
423 remote copies of rsync. By default rsync will use rsh, but you may
424 like to instead use ssh because of its high security.
426 You can also choose the remote shell program using the RSYNC_RSH
427 environment variable.
429 dit(bf(--rsync-path PATH)) Use this to specify the path to the copy of
430 rsync on the remote machine. Useful when its not in your path.
432 dit(bf(--exclude pattern)) This option allows you to selectively exclude
433 certain files from the list of files to be transferred. This is most
434 useful in combination with a recursive transfer.
436 You may use as many --exclude options on the command line as you like
437 to build up the list of files to exclude.
439 See the section of exclude patterns for information on the syntax of
442 dit(bf(--exclude-from FILE)) This option is similar to the --exclude
443 option, but instead it adds all filenames listed in the file FILE to
446 dit(bf(--include pattern)) This option tells rsync to not exclude the
447 specified pattern of filenames. This is useful as it allows you to
448 build up quite complex exclude/include rules.
450 See the section of exclude patterns for information on the syntax of
453 dit(bf(--include-from FILE)) This specifies a list of include patterns
456 dit(bf(-C, --cvs-exclude)) This is a useful shorthand for excluding a
457 broad range of files that you often don't want to transfer between
458 systems. It uses the same algorithm that CVS uses to determine if
459 a file should be ignored.
461 The exclude list is initialized to:
463 quote(RCS SCCS CVS CVS.adm RCSLOG cvslog.* tags TAGS .make.state
464 .nse_depinfo *~ #* .#* ,* *.old *.bak *.BAK *.orig *.rej .del-*
465 *.a *.o *.obj *.so *.Z *.elc *.ln core)
467 then files listed in a $HOME/.cvsignore are added to the list and any
468 files listed in the CVSIGNORE environment variable (space delimited).
470 Finally in each directory any files listed in the .cvsignore file in
471 that directory are added to the list.
473 dit(bf(--suffix SUFFIX)) This option allows you to override the default
474 backup suffix used with the -b option. The default is a ~.
476 dit(bf(--csum-length LENGTH)) By default the primary checksum used in
477 rsync is a very strong 16 byte MD4 checksum. In most cases you will
478 find that a truncated version of this checksum is quite efficient, and
479 this will decrease the size of the checksum data sent over the link,
480 making things faster.
482 You can choose the number of bytes in the truncated checksum using the
483 --csum-length option. Any value less than or equal to 16 is valid.
485 Note that if you use this option then you run the risk of ending up
486 with an incorrect target file. The risk with a value of 16 is
487 microscopic and can be safely ignored (the universe will probably end
488 before it fails) but with smaller values the risk is higher.
490 Current versions of rsync actually use an adaptive algorithm for the
491 checksum length by default, using a 16 byte file checksum to determine
492 if a 2nd pass is required with a longer block checksum. Only use this
493 option if you have read the source code and know what you are doing.
495 dit(bf(-T, --temp-dir DIR)) This option instructs rsync to use DIR as a
496 scratch directory when creating temporary copies of the files
497 transferred on the receiving side. The default behavior is to create
498 the temporary files in the receiving directory.
500 dit(bf(--compare-dest DIR)) This option instructs rsync to use DIR as an
501 additional directory to compare destination files against when doing
502 transfers. This is useful for doing transfers to a new destination while
503 leaving existing files intact, and then doing a flash-cutover when all
504 files have been successfully transfered (for example by moving directories
505 around and removing the old directory). This option increases the
506 usefulness of --partial because partially transferred files will remain in
507 the new temporary destination until they have a chance to be completed.
508 If DIR is a relative path, it is relative to the destination directory.
510 dit(bf(-z, --compress)) With this option, rsync compresses any data from
511 the source file(s) which it sends to the destination machine. This
512 option is useful on slow links. The compression method used is the
513 same method that gzip uses.
515 Note this this option typically achieves better compression ratios
516 that can be achieved by using a compressing remote shell, or a
517 compressing transport, as it takes advantage of the implicit
518 information sent for matching data blocks.
520 dit(bf(--numeric-ids)) With this option rsync will transfer numeric group
521 and user ids rather than using user and group names and mapping them
524 By default rsync will use the user name and group name to determine
525 what ownership to give files. The special uid 0 and the special group
526 0 and never mapped via user/group names even if the --numeric-ids
527 option is not specified.
529 If the source system is a daemon using chroot, or if a user or group name
530 does not exist on the destination system, then the numeric id from the
531 source system is used instead.
533 dit(bf(--timeout=TIMEOUT)) This option allows you to set a maximum IO
534 timeout in seconds. If no data is transferred for the specified time
535 then rsync will exit. The default is 0, which means no timeout.
537 dit(bf(--daemon)) This tells rsync that it is to run as a rsync
538 daemon. If standard input is a socket then rsync will assume that it
539 is being run via inetd, otherwise it will detach from the current
540 terminal and become a background daemon. The daemon will read the
541 config file (/etc/rsyncd.conf) on each connect made by a client and
542 respond to requests accordingly. See the rsyncd.conf(5) man page for more
545 dit(bf(--config FILE)) This specifies an alternate config file than
546 the default /etc/rsyncd.conf. This is only relevant when --daemon is
549 dit(bf(--port PORT)) This specifies an alternate TCP port number to use
550 rather than the default port 873.
552 dit(bf(--stats)) This tells rsync to print a verbose set of statistics
553 on the file transfer, allowing you to tell how effective the rsync
554 algorithm is for your data. This option only works in conjunction with
555 the -v (verbose) option.
557 dit(bf(--progress)) This option tells rsync to print information
558 showing the progress of the transfer. This gives a bored user
563 manpagesection(EXCLUDE PATTERNS)
565 The exclude and include patterns specified to rsync allow for flexible
566 selection of what files to transfer and what files to skip.
568 rsync build a ordered list of include/exclude options as specified on
569 the command line. When a filename is encountered rsync then checks the
570 name against each exclude/include pattern in turn. The first matching
571 pattern is acted on. If it is an exclude pattern than that file is
572 skipped. If it is an include pattern then that filename is not
573 skipped. If no matching include/exclude pattern is found then the
574 filename is not skipped.
576 The patterns themselves can take several forms. The rules are:
579 it() if the pattern starts with a / then it is matched against the
580 start of the filename, otherwise it is matched against the end of
581 the filename. Thus /foo would match a file called foo
582 at the base of the tree whereas foo would match any file
583 called foo anywhere in the tree.
585 it() if the pattern ends with a / then it will only match a
586 directory, not a file, link or device.
588 it() if the pattern contains a wildcard character from the set
589 *?[ then regular expression matching is applied using the
590 normal shell filename matching rules. Otherwise a simple string
593 it() if the pattern contains a / (not counting a trailing /) then it
594 is matched against the full filename, including any leading
595 directory. If the pattern doesn't contain a / then it is matched
596 only against the final component of the filename.
598 it() if the pattern starts with "+ " (a plus followed by a space)
599 then it is always considered a include pattern, even if specified as
600 part of an exclude option. The "+ " part is discarded before matching.
602 it() if the pattern starts with "- " (a minus followed by a space)
603 then it is always considered a exclude pattern, even if specified as
604 part of an include option. The "- " part is discarded before matching.
606 it() if the pattern is a single exclamation mark ! then the current
607 exclude list is reset, removing all previous exclude patterns.
610 The +/- rules are most useful in exclude lists, allowing you to have a
611 single exclude list that contains both include and exclude options.
613 Here are some examples:
616 it() --exclude "*.o" would exclude all filenames matching *.o
617 it() --exclude "/foo" would exclude a file in the base directory called foo
618 it() --exclude "foo/" would exclude any directory called foo
619 it() --include "*/" --include "*.c" --exclude "*" would include all
620 directories and C source files.
623 manpagesection(DIAGNOSTICS)
625 rsync occasinally produces error messages that may seem a little
626 cryptic. The one that seems to cause the most confusion is "protocol
627 version mismatch - is your shell clean?".
629 This message is usually caused by your startup scripts or remote shell
630 facility producing unwanted garbage on the stream that rsync is using
631 for its transport. The way ot diagnose this problem is to run your
632 remote shell like this:
635 rsh remotehost /bin/true > out.dat
638 then look at out.dat. If everything is working correctly then out.dat
639 should be a zero length file. You you are getting the above error from
640 rsync then you will probably find that out.dat contains some text or
641 data. Look at the contents and try to work out what is producing
642 it. The most common cause is incorrectly configued shell startup
643 scripts (such as .cshrc or .profile) that contain output statements
644 for non-interactive logins.
646 manpagesection(ENVIRONMENT VARIABLES)
650 dit(bf(CVSIGNORE)) The CVSIGNORE environment variable supplements any
651 ignore patterns in .cvsignore files. See the --cvs-exclude option for
654 dit(bf(RSYNC_RSH)) The RSYNC_RSH environment variable allows you to
655 override the default shell used as the transport for rsync. This can
656 be used instead of the -e option.
658 dit(bf(RSYNC_PASSWORD)) Setting RSYNC_PASSWORD to the required
659 password allows you to run authenticated rsync connections to a rsync
660 daemon without user intervention. Note that this does not supply a
661 password to a shell transport such as ssh.
663 dit(bf(USER) or bf(LOGNAME)) The USER or LOGNAME environment variables
664 are used to determine the default username sent to a rsync server.
666 dit(bf(HOME)) The HOME environment variable is used to find the users
667 default .cvsignore file.
683 times are transferred as unix time_t values
685 file permissions, devices etc are transferred as native numerical
688 see also the comments on the --delete option
690 Please report bugs! The rsync bug tracking system is online at
691 url(http://samba.anu.edu.au/rsync/)(http://samba.anu.edu.au/rsync/)
693 manpagesection(VERSION)
694 This man page is current for version 2.0 of rsync
696 manpagesection(CREDITS)
698 rsync is distributed under the GNU public license. See the file
701 The primary ftp site for rsync is
702 url(ftp://samba.anu.edu.au/pub/rsync)(ftp://samba.anu.edu.au/pub/rsync).
704 A WEB site is available at
705 url(http://samba.anu.edu.au/rsync/)(http://samba.anu.edu.au/rsync/)
707 We would be delighted to hear from you if you like this program.
709 This program uses the zlib compression library written by Jean-loup
710 Gailly and Mark Adler.
712 manpagesection(THANKS)
714 Thanks to Richard Brent, Brendan Mackay, Bill Waite, Stephen Rothwell
715 and David Bell for helpful suggestions and testing of rsync. I've
716 probably missed some people, my apologies if I have.
721 rsync was written by Andrew Tridgell and Paul Mackerras. They may be
722 contacted via email at tridge@samba.anu.edu.au and
723 Paul.Mackerras@cs.anu.edu.au