1 mailto(rsync-bugs@samba.org)
2 manpage(rsync)(1)(6 Nov 2006)()()
3 manpagename(rsync)(faster, flexible replacement for rcp)
6 rsync [OPTION]... SRC [SRC]... DEST
8 rsync [OPTION]... SRC [SRC]... [USER@]HOST:DEST
10 rsync [OPTION]... SRC [SRC]... [USER@]HOST::DEST
12 rsync [OPTION]... SRC [SRC]... rsync://[USER@]HOST[:PORT]/DEST
16 rsync [OPTION]... [USER@]HOST:SRC [DEST]
18 rsync [OPTION]... [USER@]HOST::SRC [DEST]
20 rsync [OPTION]... rsync://[USER@]HOST[:PORT]/SRC [DEST]
24 rsync is a program that behaves in much the same way that rcp does,
25 but has many more options and uses the rsync remote-update protocol to
26 greatly speed up file transfers when the destination file is being
29 The rsync remote-update protocol allows rsync to transfer just the
30 differences between two sets of files across the network connection, using
31 an efficient checksum-search algorithm described in the technical
32 report that accompanies this package.
34 Some of the additional features of rsync are:
37 it() support for copying links, devices, owners, groups, and permissions
38 it() exclude and exclude-from options similar to GNU tar
39 it() a CVS exclude mode for ignoring the same files that CVS would ignore
40 it() can use any transparent remote shell, including ssh or rsh
41 it() does not require super-user privileges
42 it() pipelining of file transfers to minimize latency costs
43 it() support for anonymous or authenticated rsync daemons (ideal for
47 manpagesection(GENERAL)
49 Rsync copies files either to or from a remote host, or locally on the
50 current host (it does not support copying files between two remote hosts).
52 There are two different ways for rsync to contact a remote system: using a
53 remote-shell program as the transport (such as ssh or rsh) or contacting an
54 rsync daemon directly via TCP. The remote-shell transport is used whenever
55 the source or destination path contains a single colon (:) separator after
56 a host specification. Contacting an rsync daemon directly happens when the
57 source or destination path contains a double colon (::) separator after a
58 host specification, OR when an rsync:// URL is specified (see also the
59 "USING RSYNC-DAEMON FEATURES VIA A REMOTE-SHELL CONNECTION" section for
60 an exception to this latter rule).
62 As a special case, if a single source arg is specified without a
63 destination, the files are listed in an output format similar to "ls -l".
65 As expected, if neither the source or destination path specify a remote
66 host, the copy occurs locally (see also the bf(--list-only) option).
70 See the file README for installation instructions.
72 Once installed, you can use rsync to any machine that you can access via
73 a remote shell (as well as some that you can access using the rsync
74 daemon-mode protocol). For remote transfers, a modern rsync uses ssh
75 for its communications, but it may have been configured to use a
76 different remote shell by default, such as rsh or remsh.
78 You can also specify any remote shell you like, either by using the bf(-e)
79 command line option, or by setting the RSYNC_RSH environment variable.
81 Note that rsync must be installed on both the source and destination
86 You use rsync in the same way you use rcp. You must specify a source
87 and a destination, one of which may be remote.
89 Perhaps the best way to explain the syntax is with some examples:
91 quote(tt(rsync -t *.c foo:src/))
93 This would transfer all files matching the pattern *.c from the
94 current directory to the directory src on the machine foo. If any of
95 the files already exist on the remote system then the rsync
96 remote-update protocol is used to update the file by sending only the
97 differences. See the tech report for details.
99 quote(tt(rsync -avz foo:src/bar /data/tmp))
101 This would recursively transfer all files from the directory src/bar on the
102 machine foo into the /data/tmp/bar directory on the local machine. The
103 files are transferred in "archive" mode, which ensures that symbolic
104 links, devices, attributes, permissions, ownerships, etc. are preserved
105 in the transfer. Additionally, compression will be used to reduce the
106 size of data portions of the transfer.
108 quote(tt(rsync -avz foo:src/bar/ /data/tmp))
110 A trailing slash on the source changes this behavior to avoid creating an
111 additional directory level at the destination. You can think of a trailing
112 / on a source as meaning "copy the contents of this directory" as opposed
113 to "copy the directory by name", but in both cases the attributes of the
114 containing directory are transferred to the containing directory on the
115 destination. In other words, each of the following commands copies the
116 files in the same way, including their setting of the attributes of
120 tt(rsync -av /src/foo /dest)nl()
121 tt(rsync -av /src/foo/ /dest/foo)nl()
124 Note also that host and module references don't require a trailing slash to
125 copy the contents of the default directory. For example, both of these
126 copy the remote directory's contents into "/dest":
129 tt(rsync -av host: /dest)nl()
130 tt(rsync -av host::module /dest)nl()
133 You can also use rsync in local-only mode, where both the source and
134 destination don't have a ':' in the name. In this case it behaves like
135 an improved copy command.
137 Finally, you can list all the (listable) modules available from a
138 particular rsync daemon by leaving off the module name:
140 quote(tt(rsync somehost.mydomain.com::))
142 See the following section for more details.
144 manpagesection(ADVANCED USAGE)
146 The syntax for requesting multiple files from a remote host involves using
147 quoted spaces in the SRC. Some examples:
149 quote(tt(rsync host::'modname/dir1/file1 modname/dir2/file2' /dest))
151 This would copy file1 and file2 into /dest from an rsync daemon. Each
152 additional arg must include the same "modname/" prefix as the first one,
153 and must be preceded by a single space. All other spaces are assumed
154 to be a part of the filenames.
156 quote(tt(rsync -av host:'dir1/file1 dir2/file2' /dest))
158 This would copy file1 and file2 into /dest using a remote shell. This
159 word-splitting is done by the remote shell, so if it doesn't work it means
160 that the remote shell isn't configured to split its args based on
161 whitespace (a very rare setting, but not unknown). If you need to transfer
162 a filename that contains whitespace, you'll need to either escape the
163 whitespace in a way that the remote shell will understand, or use wildcards
164 in place of the spaces. Two examples of this are:
167 tt(rsync -av host:'file\ name\ with\ spaces' /dest)nl()
168 tt(rsync -av host:file?name?with?spaces /dest)nl()
171 This latter example assumes that your shell passes through unmatched
172 wildcards. If it complains about "no match", put the name in quotes.
174 manpagesection(CONNECTING TO AN RSYNC DAEMON)
176 It is also possible to use rsync without a remote shell as the transport.
177 In this case you will directly connect to a remote rsync daemon, typically
178 using TCP port 873. (This obviously requires the daemon to be running on
179 the remote system, so refer to the STARTING AN RSYNC DAEMON TO ACCEPT
180 CONNECTIONS section below for information on that.)
182 Using rsync in this way is the same as using it with a remote shell except
186 it() you either use a double colon :: instead of a single colon to
187 separate the hostname from the path, or you use an rsync:// URL.
188 it() the first word of the "path" is actually a module name.
189 it() the remote daemon may print a message of the day when you
191 it() if you specify no path name on the remote daemon then the
192 list of accessible paths on the daemon will be shown.
193 it() if you specify no local destination then a listing of the
194 specified files on the remote daemon is provided.
195 it() you must not specify the bf(--rsh) (bf(-e)) option.
198 An example that copies all the files in a remote module named "src":
200 verb( rsync -av host::src /dest)
202 Some modules on the remote daemon may require authentication. If so,
203 you will receive a password prompt when you connect. You can avoid the
204 password prompt by setting the environment variable RSYNC_PASSWORD to
205 the password you want to use or using the bf(--password-file) option. This
206 may be useful when scripting rsync.
208 WARNING: On some systems environment variables are visible to all
209 users. On those systems using bf(--password-file) is recommended.
211 You may establish the connection via a web proxy by setting the
212 environment variable RSYNC_PROXY to a hostname:port pair pointing to
213 your web proxy. Note that your web proxy's configuration must support
214 proxy connections to port 873.
216 manpagesection(USING RSYNC-DAEMON FEATURES VIA A REMOTE-SHELL CONNECTION)
218 It is sometimes useful to use various features of an rsync daemon (such as
219 named modules) without actually allowing any new socket connections into a
220 system (other than what is already required to allow remote-shell access).
221 Rsync supports connecting to a host using a remote shell and then spawning
222 a single-use "daemon" server that expects to read its config file in the
223 home dir of the remote user. This can be useful if you want to encrypt a
224 daemon-style transfer's data, but since the daemon is started up fresh by
225 the remote user, you may not be able to use features such as chroot or
226 change the uid used by the daemon. (For another way to encrypt a daemon
227 transfer, consider using ssh to tunnel a local port to a remote machine and
228 configure a normal rsync daemon on that remote host to only allow
229 connections from "localhost".)
231 From the user's perspective, a daemon transfer via a remote-shell
232 connection uses nearly the same command-line syntax as a normal
233 rsync-daemon transfer, with the only exception being that you must
234 explicitly set the remote shell program on the command-line with the
235 bf(--rsh=COMMAND) option. (Setting the RSYNC_RSH in the environment
236 will not turn on this functionality.) For example:
238 verb( rsync -av --rsh=ssh host::module /dest)
240 If you need to specify a different remote-shell user, keep in mind that the
241 user@ prefix in front of the host is specifying the rsync-user value (for a
242 module that requires user-based authentication). This means that you must
243 give the '-l user' option to ssh when specifying the remote-shell, as in
244 this example that uses the short version of the bf(--rsh) option:
246 verb( rsync -av -e "ssh -l ssh-user" rsync-user@host::module /dest)
248 The "ssh-user" will be used at the ssh level; the "rsync-user" will be
249 used to log-in to the "module".
251 manpagesection(STARTING AN RSYNC DAEMON TO ACCEPT CONNECTIONS)
253 In order to connect to an rsync daemon, the remote system needs to have a
254 daemon already running (or it needs to have configured something like inetd
255 to spawn an rsync daemon for incoming connections on a particular port).
256 For full information on how to start a daemon that will handling incoming
257 socket connections, see the bf(rsyncd.conf)(5) man page -- that is the config
258 file for the daemon, and it contains the full details for how to run the
259 daemon (including stand-alone and inetd configurations).
261 If you're using one of the remote-shell transports for the transfer, there is
262 no need to manually start an rsync daemon.
264 manpagesection(EXAMPLES)
266 Here are some examples of how I use rsync.
268 To backup my wife's home directory, which consists of large MS Word
269 files and mail folders, I use a cron job that runs
271 quote(tt(rsync -Cavz . arvidsjaur:backup))
273 each night over a PPP connection to a duplicate directory on my machine
276 To synchronize my samba source trees I use the following Makefile
280 rsync -avuzb --exclude '*~' samba:samba/ .
282 rsync -Cavuzb . samba:samba/
285 this allows me to sync with a CVS directory at the other end of the
286 connection. I then do CVS operations on the remote machine, which saves a
287 lot of time as the remote CVS protocol isn't very efficient.
289 I mirror a directory between my "old" and "new" ftp sites with the
292 tt(rsync -az -e ssh --delete ~ftp/pub/samba nimbus:"~ftp/pub/tridge")
294 This is launched from cron every few hours.
296 manpagesection(OPTIONS SUMMARY)
298 Here is a short summary of the options available in rsync. Please refer
299 to the detailed description below for a complete description. verb(
300 -v, --verbose increase verbosity
301 -q, --quiet suppress non-error messages
302 --no-motd suppress daemon-mode MOTD (see caveat)
303 -c, --checksum skip based on checksum, not mod-time & size
304 -a, --archive archive mode; same as -rlptgoD (no -H, -A)
305 --no-OPTION turn off an implied OPTION (e.g. --no-D)
306 -r, --recursive recurse into directories
307 -R, --relative use relative path names
308 --no-implied-dirs don't send implied dirs with --relative
309 -b, --backup make backups (see --suffix & --backup-dir)
310 --backup-dir=DIR make backups into hierarchy based in DIR
311 --suffix=SUFFIX backup suffix (default ~ w/o --backup-dir)
312 -u, --update skip files that are newer on the receiver
313 --inplace update destination files in-place
314 --append append data onto shorter files
315 -d, --dirs transfer directories without recursing
316 -l, --links copy symlinks as symlinks
317 -L, --copy-links transform symlink into referent file/dir
318 --copy-unsafe-links only "unsafe" symlinks are transformed
319 --safe-links ignore symlinks that point outside the tree
320 -k, --copy-dirlinks transform symlink to dir into referent dir
321 -K, --keep-dirlinks treat symlinked dir on receiver as dir
322 -H, --hard-links preserve hard links
323 -p, --perms preserve permissions
324 -E, --executability preserve executability
325 --chmod=CHMOD affect file and/or directory permissions
326 -A, --acls preserve ACLs (implies -p)
327 -o, --owner preserve owner (super-user only)
328 -g, --group preserve group
329 --devices preserve device files (super-user only)
330 --specials preserve special files
331 -D same as --devices --specials
332 -t, --times preserve times
333 -O, --omit-dir-times omit directories when preserving times
334 --super receiver attempts super-user activities
335 -S, --sparse handle sparse files efficiently
336 -n, --dry-run show what would have been transferred
337 -W, --whole-file copy files whole (without rsync algorithm)
338 -x, --one-file-system don't cross filesystem boundaries
339 -B, --block-size=SIZE force a fixed checksum block-size
340 -e, --rsh=COMMAND specify the remote shell to use
341 --rsync-path=PROGRAM specify the rsync to run on remote machine
342 --existing skip creating new files on receiver
343 --ignore-existing skip updating files that exist on receiver
344 --remove-source-files sender removes synchronized files (non-dir)
345 --del an alias for --delete-during
346 --delete delete extraneous files from dest dirs
347 --delete-before receiver deletes before transfer (default)
348 --delete-during receiver deletes during xfer, not before
349 --delete-delay find deletions during, delete after
350 --delete-after receiver deletes after transfer, not before
351 --delete-excluded also delete excluded files from dest dirs
352 --ignore-errors delete even if there are I/O errors
353 --force force deletion of dirs even if not empty
354 --max-delete=NUM don't delete more than NUM files
355 --max-size=SIZE don't transfer any file larger than SIZE
356 --min-size=SIZE don't transfer any file smaller than SIZE
357 --partial keep partially transferred files
358 --partial-dir=DIR put a partially transferred file into DIR
359 --delay-updates put all updated files into place at end
360 -m, --prune-empty-dirs prune empty directory chains from file-list
361 --numeric-ids don't map uid/gid values by user/group name
362 --timeout=TIME set I/O timeout in seconds
363 -I, --ignore-times don't skip files that match size and time
364 --size-only skip files that match in size
365 --modify-window=NUM compare mod-times with reduced accuracy
366 -T, --temp-dir=DIR create temporary files in directory DIR
367 -y, --fuzzy find similar file for basis if no dest file
368 --compare-dest=DIR also compare received files relative to DIR
369 --copy-dest=DIR ... and include copies of unchanged files
370 --link-dest=DIR hardlink to files in DIR when unchanged
371 -z, --compress compress file data during the transfer
372 --compress-level=NUM explicitly set compression level
373 -C, --cvs-exclude auto-ignore files in the same way CVS does
374 -f, --filter=RULE add a file-filtering RULE
375 -F same as --filter='dir-merge /.rsync-filter'
376 repeated: --filter='- .rsync-filter'
377 --exclude=PATTERN exclude files matching PATTERN
378 --exclude-from=FILE read exclude patterns from FILE
379 --include=PATTERN don't exclude files matching PATTERN
380 --include-from=FILE read include patterns from FILE
381 --files-from=FILE read list of source-file names from FILE
382 -0, --from0 all *from/filter files are delimited by 0s
383 --address=ADDRESS bind address for outgoing socket to daemon
384 --port=PORT specify double-colon alternate port number
385 --sockopts=OPTIONS specify custom TCP options
386 --blocking-io use blocking I/O for the remote shell
387 --stats give some file-transfer stats
388 -8, --8-bit-output leave high-bit chars unescaped in output
389 -h, --human-readable output numbers in a human-readable format
390 --progress show progress during transfer
391 -P same as --partial --progress
392 -i, --itemize-changes output a change-summary for all updates
393 --out-format=FORMAT output updates using the specified FORMAT
394 --log-file=FILE log what we're doing to the specified FILE
395 --log-file-format=FMT log updates using the specified FMT
396 --password-file=FILE read daemon-access password from FILE
397 --list-only list the files instead of copying them
398 --bwlimit=KBPS limit I/O bandwidth; KBytes per second
399 --write-batch=FILE write a batched update to FILE
400 --only-write-batch=FILE like --write-batch but w/o updating dest
401 --read-batch=FILE read a batched update from FILE
402 --protocol=NUM force an older protocol version to be used
403 --checksum-seed=NUM set block/file checksum seed (advanced)
404 -4, --ipv4 prefer IPv4
405 -6, --ipv6 prefer IPv6
406 --version print version number
407 (-h) --help show this help (see below for -h comment))
409 Rsync can also be run as a daemon, in which case the following options are
411 --daemon run as an rsync daemon
412 --address=ADDRESS bind to the specified address
413 --bwlimit=KBPS limit I/O bandwidth; KBytes per second
414 --config=FILE specify alternate rsyncd.conf file
415 --no-detach do not detach from the parent
416 --port=PORT listen on alternate port number
417 --log-file=FILE override the "log file" setting
418 --log-file-format=FMT override the "log format" setting
419 --sockopts=OPTIONS specify custom TCP options
420 -v, --verbose increase verbosity
421 -4, --ipv4 prefer IPv4
422 -6, --ipv6 prefer IPv6
423 -h, --help show this help (if used after --daemon))
427 rsync uses the GNU long options package. Many of the command line
428 options have two variants, one short and one long. These are shown
429 below, separated by commas. Some options only have a long variant.
430 The '=' for options that take a parameter is optional; whitespace
434 dit(bf(--help)) Print a short help page describing the options
435 available in rsync and exit. For backward-compatibility with older
436 versions of rsync, the help will also be output if you use the bf(-h)
437 option without any other args.
439 dit(bf(--version)) print the rsync version number and exit.
441 dit(bf(-v, --verbose)) This option increases the amount of information you
442 are given during the transfer. By default, rsync works silently. A
443 single bf(-v) will give you information about what files are being
444 transferred and a brief summary at the end. Two bf(-v) flags will give you
445 information on what files are being skipped and slightly more
446 information at the end. More than two bf(-v) flags should only be used if
447 you are debugging rsync.
449 Note that the names of the transferred files that are output are done using
450 a default bf(--out-format) of "%n%L", which tells you just the name of the
451 file and, if the item is a link, where it points. At the single bf(-v)
452 level of verbosity, this does not mention when a file gets its attributes
453 changed. If you ask for an itemized list of changed attributes (either
454 bf(--itemize-changes) or adding "%i" to the bf(--out-format) setting), the
455 output (on the client) increases to mention all items that are changed in
456 any way. See the bf(--out-format) option for more details.
458 dit(bf(-q, --quiet)) This option decreases the amount of information you
459 are given during the transfer, notably suppressing information messages
460 from the remote server. This flag is useful when invoking rsync from
463 dit(bf(--no-motd)) This option affects the information that is output
464 by the client at the start of a daemon transfer. This suppresses the
465 message-of-the-day (MOTD) text, but it also affects the list of modules
466 that the daemon sends in response to the "rsync host::" request (due to
467 a limitation in the rsync protocol), so omit this option if you want to
468 request the list of modules from the daemon.
470 dit(bf(-I, --ignore-times)) Normally rsync will skip any files that are
471 already the same size and have the same modification time-stamp.
472 This option turns off this "quick check" behavior, causing all files to
475 dit(bf(--size-only)) Normally rsync will not transfer any files that are
476 already the same size and have the same modification time-stamp. With the
477 bf(--size-only) option, files will not be transferred if they have the same size,
478 regardless of timestamp. This is useful when starting to use rsync
479 after using another mirroring system which may not preserve timestamps
482 dit(bf(--modify-window)) When comparing two timestamps, rsync treats the
483 timestamps as being equal if they differ by no more than the modify-window
484 value. This is normally 0 (for an exact match), but you may find it useful
485 to set this to a larger value in some situations. In particular, when
486 transferring to or from an MS Windows FAT filesystem (which represents
487 times with a 2-second resolution), bf(--modify-window=1) is useful
488 (allowing times to differ by up to 1 second).
490 dit(bf(-c, --checksum)) This forces the sender to checksum em(every)
491 regular file using a 128-bit MD4 checksum. It does this during the initial
492 file-system scan as it builds the list of all available files. The receiver
493 then checksums its version of each file (if it exists and it has the same
494 size as its sender-side counterpart) in order to decide which files need to
495 be updated: files with either a changed size or a changed checksum are
496 selected for transfer. Since this whole-file checksumming of all files on
497 both sides of the connection occurs in addition to the automatic checksum
498 verifications that occur during a file's transfer, this option can be quite
501 Note that rsync always verifies that each em(transferred) file was correctly
502 reconstructed on the receiving side by checking its whole-file checksum, but
503 that automatic after-the-transfer verification has nothing to do with this
504 option's before-the-transfer "Does this file need to be updated?" check.
506 dit(bf(-a, --archive)) This is equivalent to bf(-rlptgoD). It is a quick
507 way of saying you want recursion and want to preserve almost
508 everything (with -H being a notable omission).
509 The only exception to the above equivalence is when bf(--files-from) is
510 specified, in which case bf(-r) is not implied.
512 Note that bf(-a) bf(does not preserve hardlinks), because
513 finding multiply-linked files is expensive. You must separately
516 dit(--no-OPTION) You may turn off one or more implied options by prefixing
517 the option name with "no-". Not all options may be prefixed with a "no-":
518 only options that are implied by other options (e.g. bf(--no-D),
519 bf(--no-perms)) or have different defaults in various circumstances
520 (e.g. bf(--no-whole-file), bf(--no-blocking-io), bf(--no-dirs)). You may
521 specify either the short or the long option name after the "no-" prefix
522 (e.g. bf(--no-R) is the same as bf(--no-relative)).
524 For example: if you want to use bf(-a) (bf(--archive)) but don't want
525 bf(-o) (bf(--owner)), instead of converting bf(-a) into bf(-rlptgD), you
526 could specify bf(-a --no-o) (or bf(-a --no-owner)).
528 The order of the options is important: if you specify bf(--no-r -a), the
529 bf(-r) option would end up being turned on, the opposite of bf(-a --no-r).
530 Note also that the side-effects of the bf(--files-from) option are NOT
531 positional, as it affects the default state of several options and slightly
532 changes the meaning of bf(-a) (see the bf(--files-from) option for more
535 dit(bf(-r, --recursive)) This tells rsync to copy directories
536 recursively. See also bf(--dirs) (bf(-d)).
538 Beginning with rsync 3.0.0, the recursive algorithm used is now an
539 incremental scan that uses much less memory than before and begins the
540 transfer after the scanning of the first few directories have been
541 completed. This incremental scan only affects our recursion algorithm, and
542 does not change a non-recursive transfer (e.g. when using a fully-specified
543 bf(--files-from) list). It is also only possible when both ends of the
544 transfer are at least version 3.0.0.
546 Some options require rsync to know the full file list, so these options
547 disable the incremental recursion mode. These include: bf(--delete-before),
548 bf(--delete-after), bf(--prune-empty-dirs), bf(--delay-updates), and bf(--hard-links).
549 Because of this, the default delete mode when you specify bf(--delete) is now
550 bf(--delete-during) when both ends of the connection are at least 3.0.0
551 (use bf(--del) or bf(--delete-during) to request this improved deletion mode
552 explicitly). See also the bf(--delete-delay) option that is a better choice
553 than using bf(--delete-after).
555 dit(bf(-R, --relative)) Use relative paths. This means that the full path
556 names specified on the command line are sent to the server rather than
557 just the last parts of the filenames. This is particularly useful when
558 you want to send several different directories at the same time. For
559 example, if you used this command:
561 quote(tt( rsync -av /foo/bar/baz.c remote:/tmp/))
563 ... this would create a file named baz.c in /tmp/ on the remote
564 machine. If instead you used
566 quote(tt( rsync -avR /foo/bar/baz.c remote:/tmp/))
568 then a file named /tmp/foo/bar/baz.c would be created on the remote
569 machine -- the full path name is preserved. To limit the amount of
570 path information that is sent, you have a couple options: (1) With
571 a modern rsync on the sending side (beginning with 2.6.7), you can
572 insert a dot and a slash into the source path, like this:
574 quote(tt( rsync -avR /foo/./bar/baz.c remote:/tmp/))
576 That would create /tmp/bar/baz.c on the remote machine. (Note that the
577 dot must be followed by a slash, so "/foo/." would not be abbreviated.)
578 (2) For older rsync versions, you would need to use a chdir to limit the
579 source path. For example, when pushing files:
581 quote(tt( (cd /foo; rsync -avR bar/baz.c remote:/tmp/) ))
583 (Note that the parens put the two commands into a sub-shell, so that the
584 "cd" command doesn't remain in effect for future commands.)
585 If you're pulling files, use this idiom (which doesn't work with an
589 tt( rsync -avR --rsync-path="cd /foo; rsync" \ )nl()
590 tt( remote:bar/baz.c /tmp/)
593 dit(bf(--no-implied-dirs)) This option affects the default behavior of the
594 bf(--relative) option. When it is specified, the attributes of the implied
595 directories from the source names are not included in the transfer. This
596 means that the corresponding path elements on the destination system are
597 left unchanged if they exist, and any missing implied directories are
598 created with default attributes. This even allows these implied path
599 elements to have big differences, such as being a symlink to a directory on
600 one side of the transfer, and a real directory on the other side.
602 For instance, if a command-line arg or a files-from entry told rsync to
603 transfer the file "path/foo/file", the directories "path" and "path/foo"
604 are implied when bf(--relative) is used. If "path/foo" is a symlink to
605 "bar" on the destination system, the receiving rsync would ordinarily
606 delete "path/foo", recreate it as a directory, and receive the file into
607 the new directory. With bf(--no-implied-dirs), the receiving rsync updates
608 "path/foo/file" using the existing path elements, which means that the file
609 ends up being created in "path/bar". Another way to accomplish this link
610 preservation is to use the bf(--keep-dirlinks) option (which will also
611 affect symlinks to directories in the rest of the transfer).
613 In a similar but opposite scenario, if the transfer of "path/foo/file" is
614 requested and "path/foo" is a symlink on the sending side, running without
615 bf(--no-implied-dirs) would cause rsync to transform "path/foo" on the
616 receiving side into an identical symlink, and then attempt to transfer
617 "path/foo/file", which might fail if the duplicated symlink did not point
618 to a directory on the receiving side. Another way to avoid this sending of
619 a symlink as an implied directory is to use bf(--copy-unsafe-links), or
620 bf(--copy-dirlinks) (both of which also affect symlinks in the rest of the
621 transfer -- see their descriptions for full details).
623 dit(bf(-b, --backup)) With this option, preexisting destination files are
624 renamed as each file is transferred or deleted. You can control where the
625 backup file goes and what (if any) suffix gets appended using the
626 bf(--backup-dir) and bf(--suffix) options.
628 Note that if you don't specify bf(--backup-dir), (1) the
629 bf(--omit-dir-times) option will be implied, and (2) if bf(--delete) is
630 also in effect (without bf(--delete-excluded)), rsync will add a "protect"
631 filter-rule for the backup suffix to the end of all your existing excludes
632 (e.g. bf(-f "Pp *~")). This will prevent previously backed-up files from being
633 deleted. Note that if you are supplying your own filter rules, you may
634 need to manually insert your own exclude/protect rule somewhere higher up
635 in the list so that it has a high enough priority to be effective (e.g., if
636 your rules specify a trailing inclusion/exclusion of '*', the auto-added
637 rule would never be reached).
639 dit(bf(--backup-dir=DIR)) In combination with the bf(--backup) option, this
640 tells rsync to store all backups in the specified directory on the receiving
641 side. This can be used for incremental backups. You can additionally
642 specify a backup suffix using the bf(--suffix) option
643 (otherwise the files backed up in the specified directory
644 will keep their original filenames).
646 dit(bf(--suffix=SUFFIX)) This option allows you to override the default
647 backup suffix used with the bf(--backup) (bf(-b)) option. The default suffix is a ~
648 if no -bf(-backup-dir) was specified, otherwise it is an empty string.
650 dit(bf(-u, --update)) This forces rsync to skip any files which exist on
651 the destination and have a modified time that is newer than the source
652 file. (If an existing destination file has a modify time equal to the
653 source file's, it will be updated if the sizes are different.)
655 In the current implementation of bf(--update), a difference of file format
656 between the sender and receiver is always
657 considered to be important enough for an update, no matter what date
658 is on the objects. In other words, if the source has a directory or a
659 symlink where the destination has a file, the transfer would occur
660 regardless of the timestamps. This might change in the future (feel
661 free to comment on this on the mailing list if you have an opinion).
663 dit(bf(--inplace)) This causes rsync not to create a new copy of the file
664 and then move it into place. Instead rsync will overwrite the existing
665 file, meaning that the rsync algorithm can't accomplish the full amount of
666 network reduction it might be able to otherwise (since it does not yet try
667 to sort data matches). One exception to this is if you combine the option
668 with bf(--backup), since rsync is smart enough to use the backup file as the
669 basis file for the transfer.
671 This option is useful for transfer of large files with block-based changes
672 or appended data, and also on systems that are disk bound, not network
675 The option implies bf(--partial) (since an interrupted transfer does not delete
676 the file), but conflicts with bf(--partial-dir) and bf(--delay-updates).
677 Prior to rsync 2.6.4 bf(--inplace) was also incompatible with bf(--compare-dest)
680 WARNING: The file's data will be in an inconsistent state during the
681 transfer (and possibly afterward if the transfer gets interrupted), so you
682 should not use this option to update files that are in use. Also note that
683 rsync will be unable to update a file in-place that is not writable by the
686 dit(bf(--append)) This causes rsync to update a file by appending data onto
687 the end of the file, which presumes that the data that already exists on
688 the receiving side is identical with the start of the file on the sending
689 side. If that is not true, the file will fail the checksum test, and the
690 resend will do a normal bf(--inplace) update to correct the mismatched data.
691 Only files on the receiving side that are shorter than the corresponding
692 file on the sending side (as well as new files) are sent.
693 Implies bf(--inplace), but does not conflict with bf(--sparse) (though the
694 bf(--sparse) option will be auto-disabled if a resend of the already-existing
697 dit(bf(-d, --dirs)) Tell the sending side to include any directories that
698 are encountered. Unlike bf(--recursive), a directory's contents are not copied
699 unless the directory name specified is "." or ends with a trailing slash
700 (e.g. ".", "dir/.", "dir/", etc.). Without this option or the
701 bf(--recursive) option, rsync will skip all directories it encounters (and
702 output a message to that effect for each one). If you specify both
703 bf(--dirs) and bf(--recursive), bf(--recursive) takes precedence.
705 dit(bf(-l, --links)) When symlinks are encountered, recreate the
706 symlink on the destination.
708 dit(bf(-L, --copy-links)) When symlinks are encountered, the item that
709 they point to (the referent) is copied, rather than the symlink. In older
710 versions of rsync, this option also had the side-effect of telling the
711 receiving side to follow symlinks, such as symlinks to directories. In a
712 modern rsync such as this one, you'll need to specify bf(--keep-dirlinks) (bf(-K))
713 to get this extra behavior. The only exception is when sending files to
714 an rsync that is too old to understand bf(-K) -- in that case, the bf(-L) option
715 will still have the side-effect of bf(-K) on that older receiving rsync.
717 dit(bf(--copy-unsafe-links)) This tells rsync to copy the referent of
718 symbolic links that point outside the copied tree. Absolute symlinks
719 are also treated like ordinary files, and so are any symlinks in the
720 source path itself when bf(--relative) is used. This option has no
721 additional effect if bf(--copy-links) was also specified.
723 dit(bf(--safe-links)) This tells rsync to ignore any symbolic links
724 which point outside the copied tree. All absolute symlinks are
725 also ignored. Using this option in conjunction with bf(--relative) may
726 give unexpected results.
728 dit(bf(-K, --copy-dirlinks)) This option causes the sending side to treat
729 a symlink to a directory as though it were a real directory. This is
730 useful if you don't want symlinks to non-directories to be affected, as
731 they would be using bf(--copy-links).
733 Without this option, if the sending side has replaced a directory with a
734 symlink to a directory, the receiving side will delete anything that is in
735 the way of the new symlink, including a directory hierarchy (as long as
736 bf(--force) or bf(--delete) is in effect).
738 See also bf(--keep-dirlinks) for an analogous option for the receiving
741 dit(bf(-K, --keep-dirlinks)) This option causes the receiving side to treat
742 a symlink to a directory as though it were a real directory, but only if it
743 matches a real directory from the sender. Without this option, the
744 receiver's symlink would be deleted and replaced with a real directory.
746 For example, suppose you transfer a directory "foo" that contains a file
747 "file", but "foo" is a symlink to directory "bar" on the receiver. Without
748 bf(--keep-dirlinks), the receiver deletes symlink "foo", recreates it as a
749 directory, and receives the file into the new directory. With
750 bf(--keep-dirlinks), the receiver keeps the symlink and "file" ends up in
753 See also bf(--copy-dirlinks) for an analogous option for the sending side.
755 dit(bf(-H, --hard-links)) This tells rsync to look for hard-linked files in
756 the transfer and link together the corresponding files on the receiving
757 side. Without this option, hard-linked files in the transfer are treated
758 as though they were separate files.
760 Note that rsync can only detect hard links if both parts of the link
761 are in the list of files being sent.
763 dit(bf(-p, --perms)) This option causes the receiving rsync to set the
764 destination permissions to be the same as the source permissions. (See
765 also the bf(--chmod) option for a way to modify what rsync considers to
766 be the source permissions.)
768 When this option is em(off), permissions are set as follows:
771 it() Existing files (including updated files) retain their existing
772 permissions, though the bf(--executability) option might change just
773 the execute permission for the file.
774 it() New files get their "normal" permission bits set to the source
775 file's permissions masked with the receiving directory's default
776 permissions (either the receiving process's umask, or the permissions
777 specified via the destination directory's default ACL), and
778 their special permission bits disabled except in the case where a new
779 directory inherits a setgid bit from its parent directory.
782 Thus, when bf(--perms) and bf(--executability) are both disabled,
783 rsync's behavior is the same as that of other file-copy utilities,
784 such as bf(cp)(1) and bf(tar)(1).
786 In summary: to give destination files (both old and new) the source
787 permissions, use bf(--perms). To give new files the destination-default
788 permissions (while leaving existing files unchanged), make sure that the
789 bf(--perms) option is off and use bf(--chmod=ugo=rwX) (which ensures that
790 all non-masked bits get enabled). If you'd care to make this latter
791 behavior easier to type, you could define a popt alias for it, such as
792 putting this line in the file ~/.popt (this defines the bf(-s) option,
793 and includes --no-g to use the default group of the destination dir):
795 quote(tt( rsync alias -s --no-p --no-g --chmod=ugo=rwX))
797 You could then use this new option in a command such as this one:
799 quote(tt( rsync -asv src/ dest/))
801 (Caveat: make sure that bf(-a) does not follow bf(-s), or it will re-enable
802 the "--no-*" options.)
804 The preservation of the destination's setgid bit on newly-created
805 directories when bf(--perms) is off was added in rsync 2.6.7. Older rsync
806 versions erroneously preserved the three special permission bits for
807 newly-created files when bf(--perms) was off, while overriding the
808 destination's setgid bit setting on a newly-created directory. Default ACL
809 observance was added to the ACL patch for rsync 2.6.7, so older (or
810 non-ACL-enabled) rsyncs use the umask even if default ACLs are present.
811 (Keep in mind that it is the version of the receiving rsync that affects
814 dit(bf(-E, --executability)) This option causes rsync to preserve the
815 executability (or non-executability) of regular files when bf(--perms) is
816 not enabled. A regular file is considered to be executable if at least one
817 'x' is turned on in its permissions. When an existing destination file's
818 executability differs from that of the corresponding source file, rsync
819 modifies the destination file's permissions as follows:
822 it() To make a file non-executable, rsync turns off all its 'x'
824 it() To make a file executable, rsync turns on each 'x' permission that
825 has a corresponding 'r' permission enabled.
828 If bf(--perms) is enabled, this option is ignored.
830 dit(bf(-A, --acls)) This option causes rsync to update the destination
831 ACLs to be the same as the source ACLs. This nonstandard option only
832 works if the remote rsync also supports it. bf(--acls) implies bf(--perms).
834 dit(bf(--chmod)) This option tells rsync to apply one or more
835 comma-separated "chmod" strings to the permission of the files in the
836 transfer. The resulting value is treated as though it was the permissions
837 that the sending side supplied for the file, which means that this option
838 can seem to have no effect on existing files if bf(--perms) is not enabled.
840 In addition to the normal parsing rules specified in the bf(chmod)(1)
841 manpage, you can specify an item that should only apply to a directory by
842 prefixing it with a 'D', or specify an item that should only apply to a
843 file by prefixing it with a 'F'. For example:
845 quote(--chmod=Dg+s,ug+w,Fo-w,+X)
847 It is also legal to specify multiple bf(--chmod) options, as each
848 additional option is just appended to the list of changes to make.
850 See the bf(--perms) and bf(--executability) options for how the resulting
851 permission value can be applied to the files in the transfer.
853 dit(bf(-o, --owner)) This option causes rsync to set the owner of the
854 destination file to be the same as the source file, but only if the
855 receiving rsync is being run as the super-user (see also the bf(--super)
856 option to force rsync to attempt super-user activities).
857 Without this option, the owner is set to the invoking user on the
860 The preservation of ownership will associate matching names by default, but
861 may fall back to using the ID number in some circumstances (see also the
862 bf(--numeric-ids) option for a full discussion).
864 dit(bf(-g, --group)) This option causes rsync to set the group of the
865 destination file to be the same as the source file. If the receiving
866 program is not running as the super-user (or if bf(--no-super) was
867 specified), only groups that the invoking user on the receiving side
868 is a member of will be preserved.
869 Without this option, the group is set to the default group of the invoking
870 user on the receiving side.
872 The preservation of group information will associate matching names by
873 default, but may fall back to using the ID number in some circumstances
874 (see also the bf(--numeric-ids) option for a full discussion).
876 dit(bf(--devices)) This option causes rsync to transfer character and
877 block device files to the remote system to recreate these devices.
878 This option has no effect if the receiving rsync is not run as the
879 super-user and bf(--super) is not specified.
881 dit(bf(--specials)) This option causes rsync to transfer special files
882 such as named sockets and fifos.
884 dit(bf(-D)) The bf(-D) option is equivalent to bf(--devices) bf(--specials).
886 dit(bf(-t, --times)) This tells rsync to transfer modification times along
887 with the files and update them on the remote system. Note that if this
888 option is not used, the optimization that excludes files that have not been
889 modified cannot be effective; in other words, a missing bf(-t) or bf(-a) will
890 cause the next transfer to behave as if it used bf(-I), causing all files to be
891 updated (though the rsync algorithm will make the update fairly efficient
892 if the files haven't actually changed, you're much better off using bf(-t)).
894 dit(bf(-O, --omit-dir-times)) This tells rsync to omit directories when
895 it is preserving modification times (see bf(--times)). If NFS is sharing
896 the directories on the receiving side, it is a good idea to use bf(-O).
897 This option is inferred if you use bf(--backup) without bf(--backup-dir).
899 dit(bf(--super)) This tells the receiving side to attempt super-user
900 activities even if the receiving rsync wasn't run by the super-user. These
901 activities include: preserving users via the bf(--owner) option, preserving
902 all groups (not just the current user's groups) via the bf(--groups)
903 option, and copying devices via the bf(--devices) option. This is useful
904 for systems that allow such activities without being the super-user, and
905 also for ensuring that you will get errors if the receiving side isn't
906 being running as the super-user. To turn off super-user activities, the
907 super-user can use bf(--no-super).
909 dit(bf(-S, --sparse)) Try to handle sparse files efficiently so they take
910 up less space on the destination. Conflicts with bf(--inplace) because it's
911 not possible to overwrite data in a sparse fashion.
913 NOTE: Don't use this option when the destination is a Solaris "tmpfs"
914 filesystem. It doesn't seem to handle seeks over null regions
915 correctly and ends up corrupting the files.
917 dit(bf(-n, --dry-run)) This tells rsync to not do any file transfers,
918 instead it will just report the actions it would have taken.
920 dit(bf(-W, --whole-file)) With this option the incremental rsync algorithm
921 is not used and the whole file is sent as-is instead. The transfer may be
922 faster if this option is used when the bandwidth between the source and
923 destination machines is higher than the bandwidth to disk (especially when the
924 "disk" is actually a networked filesystem). This is the default when both
925 the source and destination are specified as local paths.
927 dit(bf(-x, --one-file-system)) This tells rsync to avoid crossing a
928 filesystem boundary when recursing. This does not limit the user's ability
929 to specify items to copy from multiple filesystems, just rsync's recursion
930 through the hierarchy of each directory that the user specified, and also
931 the analogous recursion on the receiving side during deletion. Also keep
932 in mind that rsync treats a "bind" mount to the same device as being on the
935 If this option is repeated, rsync omits all mount-point directories from
936 the copy. Otherwise, it includes an empty directory at each mount-point it
937 encounters (using the attributes of the mounted directory because those of
938 the underlying mount-point directory are inaccessible).
940 If rsync has been told to collapse symlinks (via bf(--copy-links) or
941 bf(--copy-unsafe-links)), a symlink to a directory on another device is
942 treated like a mount-point. Symlinks to non-directories are unaffected
945 dit(bf(--existing, --ignore-non-existing)) This tells rsync to skip
946 creating files (including directories) that do not exist
947 yet on the destination. If this option is
948 combined with the bf(--ignore-existing) option, no files will be updated
949 (which can be useful if all you want to do is to delete extraneous files).
951 dit(bf(--ignore-existing)) This tells rsync to skip updating files that
952 already exist on the destination (this does em(not) ignore existing
953 directories, or nothing would get done). See also bf(--existing).
955 dit(bf(--remove-source-files)) This tells rsync to remove from the sending
956 side the files (meaning non-directories) that are a part of the transfer
957 and have been successfully duplicated on the receiving side.
959 dit(bf(--delete)) This tells rsync to delete extraneous files from the
960 receiving side (ones that aren't on the sending side), but only for the
961 directories that are being synchronized. You must have asked rsync to
962 send the whole directory (e.g. "dir" or "dir/") without using a wildcard
963 for the directory's contents (e.g. "dir/*") since the wildcard is expanded
964 by the shell and rsync thus gets a request to transfer individual files, not
965 the files' parent directory. Files that are excluded from transfer are
966 also excluded from being deleted unless you use the bf(--delete-excluded)
967 option or mark the rules as only matching on the sending side (see the
968 include/exclude modifiers in the FILTER RULES section).
970 Prior to rsync 2.6.7, this option would have no effect unless bf(--recursive)
971 was enabled. Beginning with 2.6.7, deletions will also occur when bf(--dirs)
972 (bf(-d)) is enabled, but only for directories whose contents are being copied.
974 This option can be dangerous if used incorrectly! It is a very good idea
975 to run first using the bf(--dry-run) option (bf(-n)) to see what files would be
976 deleted to make sure important files aren't listed.
978 If the sending side detects any I/O errors, then the deletion of any
979 files at the destination will be automatically disabled. This is to
980 prevent temporary filesystem failures (such as NFS errors) on the
981 sending side causing a massive deletion of files on the
982 destination. You can override this with the bf(--ignore-errors) option.
984 The bf(--delete) option may be combined with one of the --delete-WHEN options
985 without conflict, as well as bf(--delete-excluded). However, if none of the
986 --delete-WHEN options are specified, rsync will choose the
987 bf(--delete-during) algorithm when talking to an rsync 3.0.0 or newer, and
988 the bf(--delete-before) algorithm when talking to an older rsync. See also
989 bf(--delete-delay) and bf(--delete-after).
991 dit(bf(--delete-before)) Request that the file-deletions on the receiving
992 side be done before the transfer starts.
993 See bf(--delete) (which is implied) for more details on file-deletion.
995 Deleting before the transfer is helpful if the filesystem is tight for space
996 and removing extraneous files would help to make the transfer possible.
997 However, it does introduce a delay before the start of the transfer,
998 and this delay might cause the transfer to timeout (if bf(--timeout) was
999 specified). It also forces rsync to use the old, non-incremental recursion
1000 algorithm that requires rsync to scan all the files in the transfer into
1001 memory at once (see bf(--recursive)).
1003 dit(bf(--delete-during, --del)) Request that the file-deletions on the
1004 receiving side be done incrementally as the transfer happens. This is
1005 a faster method than choosing the before- or after-transfer algorithm,
1006 but it is only supported beginning with rsync version 2.6.4.
1007 See bf(--delete) (which is implied) for more details on file-deletion.
1009 dit(bf(--delete-delay)) Request that the file-deletions on the receiving
1010 side be computed during the transfer, and then removed after the transfer
1011 completes. If the number of removed files overflows an internal buffer, a
1012 temporary file will be created on the receiving side to hold the names (it
1013 is removed while open, so you shouldn't see it during the transfer). If
1014 the creation of the temporary file fails, rsync will try to fall back to
1015 using bf(--delete-after) (which it cannot do if bf(--recursive) is doing an
1018 dit(bf(--delete-after)) Request that the file-deletions on the receiving
1019 side be done after the transfer has completed. This is useful if you
1020 are sending new per-directory merge files as a part of the transfer and
1021 you want their exclusions to take effect for the delete phase of the
1022 current transfer. It also forces rsync to use the old, non-incremental
1023 recursion algorithm that requires rsync to scan all the files in the
1024 transfer into memory at once (see bf(--recursive)).
1025 See bf(--delete) (which is implied) for more details on file-deletion.
1027 dit(bf(--delete-excluded)) In addition to deleting the files on the
1028 receiving side that are not on the sending side, this tells rsync to also
1029 delete any files on the receiving side that are excluded (see bf(--exclude)).
1030 See the FILTER RULES section for a way to make individual exclusions behave
1031 this way on the receiver, and for a way to protect files from
1032 bf(--delete-excluded).
1033 See bf(--delete) (which is implied) for more details on file-deletion.
1035 dit(bf(--ignore-errors)) Tells bf(--delete) to go ahead and delete files
1036 even when there are I/O errors.
1038 dit(bf(--force)) This option tells rsync to delete a non-empty directory
1039 when it is to be replaced by a non-directory. This is only relevant if
1040 deletions are not active (see bf(--delete) for details).
1042 Note for older rsync versions: bf(--force) used to still be required when
1043 using bf(--delete-after), and it used to be non-functional unless the
1044 bf(--recursive) option was also enabled.
1046 dit(bf(--max-delete=NUM)) This tells rsync not to delete more than NUM
1047 files or directories.
1048 Beginning with version 3.0.0, you may specify bf(--max-delete=0) to
1049 be warned about any extraneous files in the destination, but be very
1050 careful to never specify a 0 value to an older rsync client, or the
1051 option will be silently ignored. (A 3.0.0 client will die with an
1052 error if the remote rsync is not new enough to handle the situation.)
1053 This is useful when mirroring very large trees to prevent disasters.
1055 dit(bf(--max-size=SIZE)) This tells rsync to avoid transferring any
1056 file that is larger than the specified SIZE. The SIZE value can be
1057 suffixed with a string to indicate a size multiplier, and
1058 may be a fractional value (e.g. "bf(--max-size=1.5m)").
1060 The suffixes are as follows: "K" (or "KiB") is a kibibyte (1024),
1061 "M" (or "MiB") is a mebibyte (1024*1024), and "G" (or "GiB") is a
1062 gibibyte (1024*1024*1024).
1063 If you want the multiplier to be 1000 instead of 1024, use "KB",
1064 "MB", or "GB". (Note: lower-case is also accepted for all values.)
1065 Finally, if the suffix ends in either "+1" or "-1", the value will
1066 be offset by one byte in the indicated direction.
1068 Examples: --max-size=1.5mb-1 is 1499999 bytes, and --max-size=2g+1 is
1071 dit(bf(--min-size=SIZE)) This tells rsync to avoid transferring any
1072 file that is smaller than the specified SIZE, which can help in not
1073 transferring small, junk files.
1074 See the bf(--max-size) option for a description of SIZE.
1076 dit(bf(-B, --block-size=BLOCKSIZE)) This forces the block size used in
1077 the rsync algorithm to a fixed value. It is normally selected based on
1078 the size of each file being updated. See the technical report for details.
1080 dit(bf(-e, --rsh=COMMAND)) This option allows you to choose an alternative
1081 remote shell program to use for communication between the local and
1082 remote copies of rsync. Typically, rsync is configured to use ssh by
1083 default, but you may prefer to use rsh on a local network.
1085 If this option is used with bf([user@]host::module/path), then the
1086 remote shell em(COMMAND) will be used to run an rsync daemon on the
1087 remote host, and all data will be transmitted through that remote
1088 shell connection, rather than through a direct socket connection to a
1089 running rsync daemon on the remote host. See the section "USING
1090 RSYNC-DAEMON FEATURES VIA A REMOTE-SHELL CONNECTION" above.
1092 Command-line arguments are permitted in COMMAND provided that COMMAND is
1093 presented to rsync as a single argument. You must use spaces (not tabs
1094 or other whitespace) to separate the command and args from each other,
1095 and you can use single- and/or double-quotes to preserve spaces in an
1096 argument (but not backslashes). Note that doubling a single-quote
1097 inside a single-quoted string gives you a single-quote; likewise for
1098 double-quotes (though you need to pay attention to which quotes your
1099 shell is parsing and which quotes rsync is parsing). Some examples:
1102 tt( -e 'ssh -p 2234')nl()
1103 tt( -e 'ssh -o "ProxyCommand nohup ssh firewall nc -w1 %h %p"')nl()
1106 (Note that ssh users can alternately customize site-specific connect
1107 options in their .ssh/config file.)
1109 You can also choose the remote shell program using the RSYNC_RSH
1110 environment variable, which accepts the same range of values as bf(-e).
1112 See also the bf(--blocking-io) option which is affected by this option.
1114 dit(bf(--rsync-path=PROGRAM)) Use this to specify what program is to be run
1115 on the remote machine to start-up rsync. Often used when rsync is not in
1116 the default remote-shell's path (e.g. --rsync-path=/usr/local/bin/rsync).
1117 Note that PROGRAM is run with the help of a shell, so it can be any
1118 program, script, or command sequence you'd care to run, so long as it does
1119 not corrupt the standard-in & standard-out that rsync is using to
1122 One tricky example is to set a different default directory on the remote
1123 machine for use with the bf(--relative) option. For instance:
1125 quote(tt( rsync -avR --rsync-path="cd /a/b && rsync" host:c/d /e/))
1127 dit(bf(-C, --cvs-exclude)) This is a useful shorthand for excluding a
1128 broad range of files that you often don't want to transfer between
1129 systems. It uses a similar algorithm to CVS to determine if
1130 a file should be ignored.
1132 The exclude list is initialized to exclude the following items (these
1133 initial items are marked as perishable -- see the FILTER RULES section):
1135 quote(quote(tt(RCS SCCS CVS CVS.adm RCSLOG cvslog.* tags TAGS .make.state
1136 .nse_depinfo *~ #* .#* ,* _$* *$ *.old *.bak *.BAK *.orig *.rej
1137 .del-* *.a *.olb *.o *.obj *.so *.exe *.Z *.elc *.ln core .svn/ .bzr/)))
1139 then, files listed in a $HOME/.cvsignore are added to the list and any
1140 files listed in the CVSIGNORE environment variable (all cvsignore names
1141 are delimited by whitespace).
1143 Finally, any file is ignored if it is in the same directory as a
1144 .cvsignore file and matches one of the patterns listed therein. Unlike
1145 rsync's filter/exclude files, these patterns are split on whitespace.
1146 See the bf(cvs)(1) manual for more information.
1148 If you're combining bf(-C) with your own bf(--filter) rules, you should
1149 note that these CVS excludes are appended at the end of your own rules,
1150 regardless of where the bf(-C) was placed on the command-line. This makes them
1151 a lower priority than any rules you specified explicitly. If you want to
1152 control where these CVS excludes get inserted into your filter rules, you
1153 should omit the bf(-C) as a command-line option and use a combination of
1154 bf(--filter=:C) and bf(--filter=-C) (either on your command-line or by
1155 putting the ":C" and "-C" rules into a filter file with your other rules).
1156 The first option turns on the per-directory scanning for the .cvsignore
1157 file. The second option does a one-time import of the CVS excludes
1160 dit(bf(-f, --filter=RULE)) This option allows you to add rules to selectively
1161 exclude certain files from the list of files to be transferred. This is
1162 most useful in combination with a recursive transfer.
1164 You may use as many bf(--filter) options on the command line as you like
1165 to build up the list of files to exclude.
1167 See the FILTER RULES section for detailed information on this option.
1169 dit(bf(-F)) The bf(-F) option is a shorthand for adding two bf(--filter) rules to
1170 your command. The first time it is used is a shorthand for this rule:
1172 quote(tt( --filter='dir-merge /.rsync-filter'))
1174 This tells rsync to look for per-directory .rsync-filter files that have
1175 been sprinkled through the hierarchy and use their rules to filter the
1176 files in the transfer. If bf(-F) is repeated, it is a shorthand for this
1179 quote(tt( --filter='exclude .rsync-filter'))
1181 This filters out the .rsync-filter files themselves from the transfer.
1183 See the FILTER RULES section for detailed information on how these options
1186 dit(bf(--exclude=PATTERN)) This option is a simplified form of the
1187 bf(--filter) option that defaults to an exclude rule and does not allow
1188 the full rule-parsing syntax of normal filter rules.
1190 See the FILTER RULES section for detailed information on this option.
1192 dit(bf(--exclude-from=FILE)) This option is related to the bf(--exclude)
1193 option, but it specifies a FILE that contains exclude patterns (one per line).
1194 Blank lines in the file and lines starting with ';' or '#' are ignored.
1195 If em(FILE) is bf(-), the list will be read from standard input.
1197 dit(bf(--include=PATTERN)) This option is a simplified form of the
1198 bf(--filter) option that defaults to an include rule and does not allow
1199 the full rule-parsing syntax of normal filter rules.
1201 See the FILTER RULES section for detailed information on this option.
1203 dit(bf(--include-from=FILE)) This option is related to the bf(--include)
1204 option, but it specifies a FILE that contains include patterns (one per line).
1205 Blank lines in the file and lines starting with ';' or '#' are ignored.
1206 If em(FILE) is bf(-), the list will be read from standard input.
1208 dit(bf(--files-from=FILE)) Using this option allows you to specify the
1209 exact list of files to transfer (as read from the specified FILE or bf(-)
1210 for standard input). It also tweaks the default behavior of rsync to make
1211 transferring just the specified files and directories easier:
1214 it() The bf(--relative) (bf(-R)) option is implied, which preserves the path
1215 information that is specified for each item in the file (use
1216 bf(--no-relative) or bf(--no-R) if you want to turn that off).
1217 it() The bf(--dirs) (bf(-d)) option is implied, which will create directories
1218 specified in the list on the destination rather than noisily skipping
1219 them (use bf(--no-dirs) or bf(--no-d) if you want to turn that off).
1220 it() The bf(--archive) (bf(-a)) option's behavior does not imply bf(--recursive)
1221 (bf(-r)), so specify it explicitly, if you want it.
1222 it() These side-effects change the default state of rsync, so the position
1223 of the bf(--files-from) option on the command-line has no bearing on how
1224 other options are parsed (e.g. bf(-a) works the same before or after
1225 bf(--files-from), as does bf(--no-R) and all other options).
1228 The file names that are read from the FILE are all relative to the
1229 source dir -- any leading slashes are removed and no ".." references are
1230 allowed to go higher than the source dir. For example, take this
1233 quote(tt( rsync -a --files-from=/tmp/foo /usr remote:/backup))
1235 If /tmp/foo contains the string "bin" (or even "/bin"), the /usr/bin
1236 directory will be created as /backup/bin on the remote host. If it
1237 contains "bin/" (note the trailing slash), the immediate contents of
1238 the directory would also be sent (without needing to be explicitly
1239 mentioned in the file -- this began in version 2.6.4). In both cases,
1240 if the bf(-r) option was enabled, that dir's entire hierarchy would
1241 also be transferred (keep in mind that bf(-r) needs to be specified
1242 explicitly with bf(--files-from), since it is not implied by bf(-a)).
1244 that the effect of the (enabled by default) bf(--relative) option is to
1245 duplicate only the path info that is read from the file -- it does not
1246 force the duplication of the source-spec path (/usr in this case).
1248 In addition, the bf(--files-from) file can be read from the remote host
1249 instead of the local host if you specify a "host:" in front of the file
1250 (the host must match one end of the transfer). As a short-cut, you can
1251 specify just a prefix of ":" to mean "use the remote end of the
1252 transfer". For example:
1254 quote(tt( rsync -a --files-from=:/path/file-list src:/ /tmp/copy))
1256 This would copy all the files specified in the /path/file-list file that
1257 was located on the remote "src" host.
1259 dit(bf(-0, --from0)) This tells rsync that the rules/filenames it reads from a
1260 file are terminated by a null ('\0') character, not a NL, CR, or CR+LF.
1261 This affects bf(--exclude-from), bf(--include-from), bf(--files-from), and any
1262 merged files specified in a bf(--filter) rule.
1263 It does not affect bf(--cvs-exclude) (since all names read from a .cvsignore
1264 file are split on whitespace).
1266 dit(bf(-T, --temp-dir=DIR)) This option instructs rsync to use DIR as a
1267 scratch directory when creating temporary copies of the files transferred
1268 on the receiving side. The default behavior is to create each temporary
1269 file in the same directory as the associated destination file.
1271 This option is most often used when the receiving disk partition does not
1272 have enough free space to hold a copy of the largest file in the transfer.
1273 In this case (i.e. when the scratch directory in on a different disk
1274 partition), rsync will not be able to rename each received temporary file
1275 over the top of the associated destination file, but instead must copy it
1276 into place. Rsync does this by copying the file over the top of the
1277 destination file, which means that the destination file will contain
1278 truncated data during this copy. If this were not done this way (even if
1279 the destination file were first removed, the data locally copied to a
1280 temporary file in the destination directory, and then renamed into place)
1281 it would be possible for the old file to continue taking up disk space (if
1282 someone had it open), and thus there might not be enough room to fit the
1283 new version on the disk at the same time.
1285 If you are using this option for reasons other than a shortage of disk
1286 space, you may wish to combine it with the bf(--delay-updates) option,
1287 which will ensure that all copied files get put into subdirectories in the
1288 destination hierarchy, awaiting the end of the transfer. If you don't
1289 have enough room to duplicate all the arriving files on the destination
1290 partition, another way to tell rsync that you aren't overly concerned
1291 about disk space is to use the bf(--partial-dir) option with a relative
1292 path; because this tells rsync that it is OK to stash off a copy of a
1293 single file in a subdir in the destination hierarchy, rsync will use the
1294 partial-dir as a staging area to bring over the copied file, and then
1295 rename it into place from there. (Specifying a bf(--partial-dir) with
1296 an absolute path does not have this side-effect.)
1298 dit(bf(-y, --fuzzy)) This option tells rsync that it should look for a
1299 basis file for any destination file that is missing. The current algorithm
1300 looks in the same directory as the destination file for either a file that
1301 has an identical size and modified-time, or a similarly-named file. If
1302 found, rsync uses the fuzzy basis file to try to speed up the transfer.
1304 Note that the use of the bf(--delete) option might get rid of any potential
1305 fuzzy-match files, so either use bf(--delete-after) or specify some
1306 filename exclusions if you need to prevent this.
1308 dit(bf(--compare-dest=DIR)) This option instructs rsync to use em(DIR) on
1309 the destination machine as an additional hierarchy to compare destination
1310 files against doing transfers (if the files are missing in the destination
1311 directory). If a file is found in em(DIR) that is identical to the
1312 sender's file, the file will NOT be transferred to the destination
1313 directory. This is useful for creating a sparse backup of just files that
1314 have changed from an earlier backup.
1316 Beginning in version 2.6.4, multiple bf(--compare-dest) directories may be
1317 provided, which will cause rsync to search the list in the order specified
1319 If a match is found that differs only in attributes, a local copy is made
1320 and the attributes updated.
1321 If a match is not found, a basis file from one of the em(DIR)s will be
1322 selected to try to speed up the transfer.
1324 If em(DIR) is a relative path, it is relative to the destination directory.
1325 See also bf(--copy-dest) and bf(--link-dest).
1327 dit(bf(--copy-dest=DIR)) This option behaves like bf(--compare-dest), but
1328 rsync will also copy unchanged files found in em(DIR) to the destination
1329 directory using a local copy.
1330 This is useful for doing transfers to a new destination while leaving
1331 existing files intact, and then doing a flash-cutover when all files have
1332 been successfully transferred.
1334 Multiple bf(--copy-dest) directories may be provided, which will cause
1335 rsync to search the list in the order specified for an unchanged file.
1336 If a match is not found, a basis file from one of the em(DIR)s will be
1337 selected to try to speed up the transfer.
1339 If em(DIR) is a relative path, it is relative to the destination directory.
1340 See also bf(--compare-dest) and bf(--link-dest).
1342 dit(bf(--link-dest=DIR)) This option behaves like bf(--copy-dest), but
1343 unchanged files are hard linked from em(DIR) to the destination directory.
1344 The files must be identical in all preserved attributes (e.g. permissions,
1345 possibly ownership) in order for the files to be linked together.
1348 quote(tt( rsync -av --link-dest=$PWD/prior_dir host:src_dir/ new_dir/))
1350 Beginning in version 2.6.4, multiple bf(--link-dest) directories may be
1351 provided, which will cause rsync to search the list in the order specified
1353 If a match is found that differs only in attributes, a local copy is made
1354 and the attributes updated.
1355 If a match is not found, a basis file from one of the em(DIR)s will be
1356 selected to try to speed up the transfer.
1358 This option works best when copying into an empty destination hierarchy, as
1359 rsync treats existing files as definitive (so it never looks in the link-dest
1360 dirs when a destination file already exists), and as malleable (so it might
1361 change the attributes of a destination file, which affects all the hard-linked
1364 Note that if you combine this option with bf(--ignore-times), rsync will not
1365 link any files together because it only links identical files together as a
1366 substitute for transferring the file, never as an additional check after the
1369 If em(DIR) is a relative path, it is relative to the destination directory.
1370 See also bf(--compare-dest) and bf(--copy-dest).
1372 Note that rsync versions prior to 2.6.1 had a bug that could prevent
1373 bf(--link-dest) from working properly for a non-super-user when bf(-o) was
1374 specified (or implied by bf(-a)). You can work-around this bug by avoiding
1375 the bf(-o) option when sending to an old rsync.
1377 dit(bf(-z, --compress)) With this option, rsync compresses the file data
1378 as it is sent to the destination machine, which reduces the amount of data
1379 being transmitted -- something that is useful over a slow connection.
1381 Note that this option typically achieves better compression ratios than can
1382 be achieved by using a compressing remote shell or a compressing transport
1383 because it takes advantage of the implicit information in the matching data
1384 blocks that are not explicitly sent over the connection.
1386 dit(bf(--compress-level=NUM)) Explicitly set the compression level to use
1387 (see bf(--compress)) instead of letting it default. If NUM is non-zero,
1388 the bf(--compress) option is implied.
1390 dit(bf(--numeric-ids)) With this option rsync will transfer numeric group
1391 and user IDs rather than using user and group names and mapping them
1394 By default rsync will use the username and groupname to determine
1395 what ownership to give files. The special uid 0 and the special group
1396 0 are never mapped via user/group names even if the bf(--numeric-ids)
1397 option is not specified.
1399 If a user or group has no name on the source system or it has no match
1400 on the destination system, then the numeric ID
1401 from the source system is used instead. See also the comments on the
1402 "use chroot" setting in the rsyncd.conf manpage for information on how
1403 the chroot setting affects rsync's ability to look up the names of the
1404 users and groups and what you can do about it.
1406 dit(bf(--timeout=TIMEOUT)) This option allows you to set a maximum I/O
1407 timeout in seconds. If no data is transferred for the specified time
1408 then rsync will exit. The default is 0, which means no timeout.
1410 dit(bf(--address)) By default rsync will bind to the wildcard address when
1411 connecting to an rsync daemon. The bf(--address) option allows you to
1412 specify a specific IP address (or hostname) to bind to. See also this
1413 option in the bf(--daemon) mode section.
1415 dit(bf(--port=PORT)) This specifies an alternate TCP port number to use
1416 rather than the default of 873. This is only needed if you are using the
1417 double-colon (::) syntax to connect with an rsync daemon (since the URL
1418 syntax has a way to specify the port as a part of the URL). See also this
1419 option in the bf(--daemon) mode section.
1421 dit(bf(--sockopts)) This option can provide endless fun for people
1422 who like to tune their systems to the utmost degree. You can set all
1423 sorts of socket options which may make transfers faster (or
1424 slower!). Read the man page for the code(setsockopt()) system call for
1425 details on some of the options you may be able to set. By default no
1426 special socket options are set. This only affects direct socket
1427 connections to a remote rsync daemon. This option also exists in the
1428 bf(--daemon) mode section.
1430 dit(bf(--blocking-io)) This tells rsync to use blocking I/O when launching
1431 a remote shell transport. If the remote shell is either rsh or remsh,
1432 rsync defaults to using
1433 blocking I/O, otherwise it defaults to using non-blocking I/O. (Note that
1434 ssh prefers non-blocking I/O.)
1436 dit(bf(-i, --itemize-changes)) Requests a simple itemized list of the
1437 changes that are being made to each file, including attribute changes.
1438 This is exactly the same as specifying bf(--out-format='%i %n%L').
1439 If you repeat the option, unchanged files will also be output, but only
1440 if the receiving rsync is at least version 2.6.7 (you can use bf(-vv)
1441 with older versions of rsync, but that also turns on the output of other
1444 The "%i" escape has a cryptic output that is 11 letters long. The general
1445 format is like the string bf(YXcstpoguax), where bf(Y) is replaced by the
1446 type of update being done, bf(X) is replaced by the file-type, and the
1447 other letters represent attributes that may be output if they are being
1450 The update types that replace the bf(Y) are as follows:
1453 it() A bf(<) means that a file is being transferred to the remote host
1455 it() A bf(>) means that a file is being transferred to the local host
1457 it() A bf(c) means that a local change/creation is occurring for the item
1458 (such as the creation of a directory or the changing of a symlink, etc.).
1459 it() A bf(h) means that the item is a hard link to another item (requires
1461 it() A bf(.) means that the item is not being updated (though it might
1462 have attributes that are being modified).
1465 The file-types that replace the bf(X) are: bf(f) for a file, a bf(d) for a
1466 directory, an bf(L) for a symlink, a bf(D) for a device, and a bf(S) for a
1467 special file (e.g. named sockets and fifos).
1469 The other letters in the string above are the actual letters that
1470 will be output if the associated attribute for the item is being updated or
1471 a "." for no change. Three exceptions to this are: (1) a newly created
1472 item replaces each letter with a "+", (2) an identical item replaces the
1473 dots with spaces, and (3) an unknown attribute replaces each letter with
1474 a "?" (this can happen when talking to an older rsync).
1476 The attribute that is associated with each letter is as follows:
1479 it() A bf(c) means the checksum of the file is different and will be
1480 updated by the file transfer (requires bf(--checksum)).
1481 it() A bf(s) means the size of the file is different and will be updated
1482 by the file transfer.
1483 it() A bf(t) means the modification time is different and is being updated
1484 to the sender's value (requires bf(--times)). An alternate value of bf(T)
1485 means that the time will be set to the transfer time, which happens
1486 anytime a symlink is transferred, or when a file or device is transferred
1487 without bf(--times).
1488 it() A bf(p) means the permissions are different and are being updated to
1489 the sender's value (requires bf(--perms)).
1490 it() An bf(o) means the owner is different and is being updated to the
1491 sender's value (requires bf(--owner) and super-user privileges).
1492 it() A bf(g) means the group is different and is being updated to the
1493 sender's value (requires bf(--group) and the authority to set the group).
1494 it() The bf(u) slot is reserved for reporting update (access) time changes
1495 (a feature that is not yet released).
1496 it() The bf(a) means that the ACL information changed.
1497 it() The bf(x) slot is reserved for reporting extended attribute changes
1498 (a feature that is not yet released).
1501 One other output is possible: when deleting files, the "%i" will output
1502 the string "*deleting" for each item that is being removed (assuming that
1503 you are talking to a recent enough rsync that it logs deletions instead of
1504 outputting them as a verbose message).
1506 dit(bf(--out-format=FORMAT)) This allows you to specify exactly what the
1507 rsync client outputs to the user on a per-update basis. The format is a text
1508 string containing embedded single-character escape sequences prefixed with
1509 a percent (%) character. For a list of the possible escape characters, see
1510 the "log format" setting in the rsyncd.conf manpage.
1512 Specifying this option will mention each file, dir, etc. that gets updated
1513 in a significant way (a transferred file, a recreated symlink/device, or a
1514 touched directory). In addition, if the itemize-changes escape (%i) is
1515 included in the string, the logging of names increases to mention any
1516 item that is changed in any way (as long as the receiving side is at least
1517 2.6.4). See the bf(--itemize-changes) option for a description of the
1520 The bf(--verbose) option implies a format of "%n%L", but you can use
1521 bf(--out-format) without bf(--verbose) if you like, or you can override
1522 the format of its per-file output using this option.
1524 Rsync will output the out-format string prior to a file's transfer unless
1525 one of the transfer-statistic escapes is requested, in which case the
1526 logging is done at the end of the file's transfer. When this late logging
1527 is in effect and bf(--progress) is also specified, rsync will also output
1528 the name of the file being transferred prior to its progress information
1529 (followed, of course, by the out-format output).
1531 dit(bf(--log-file=FILE)) This option causes rsync to log what it is doing
1532 to a file. This is similar to the logging that a daemon does, but can be
1533 requested for the client side and/or the server side of a non-daemon
1534 transfer. If specified as a client option, transfer logging will be
1535 enabled with a default format of "%i %n%L". See the bf(--log-file-format)
1536 option if you wish to override this.
1538 Here's a example command that requests the remote side to log what is
1541 verb( rsync -av --rsync-path="rsync --log-file=/tmp/rlog" src/ dest/)
1543 This is very useful if you need to debug why a connection is closing
1546 dit(bf(--log-file-format=FORMAT)) This allows you to specify exactly what
1547 per-update logging is put into the file specified by the bf(--log-file) option
1548 (which must also be specified for this option to have any effect). If you
1549 specify an empty string, updated files will not be mentioned in the log file.
1550 For a list of the possible escape characters, see the "log format" setting
1551 in the rsyncd.conf manpage.
1553 dit(bf(--stats)) This tells rsync to print a verbose set of statistics
1554 on the file transfer, allowing you to tell how effective the rsync
1555 algorithm is for your data.
1557 The current statistics are as follows: quote(itemization(
1558 it() bf(Number of files) is the count of all "files" (in the generic
1559 sense), which includes directories, symlinks, etc.
1560 it() bf(Number of files transferred) is the count of normal files that
1561 were updated via the rsync algorithm, which does not include created
1562 dirs, symlinks, etc.
1563 it() bf(Total file size) is the total sum of all file sizes in the transfer.
1564 This does not count any size for directories or special files, but does
1565 include the size of symlinks.
1566 it() bf(Total transferred file size) is the total sum of all files sizes
1567 for just the transferred files.
1568 it() bf(Literal data) is how much unmatched file-update data we had to
1569 send to the receiver for it to recreate the updated files.
1570 it() bf(Matched data) is how much data the receiver got locally when
1571 recreating the updated files.
1572 it() bf(File list size) is how big the file-list data was when the sender
1573 sent it to the receiver. This is smaller than the in-memory size for the
1574 file list due to some compressing of duplicated data when rsync sends the
1576 it() bf(File list generation time) is the number of seconds that the
1577 sender spent creating the file list. This requires a modern rsync on the
1578 sending side for this to be present.
1579 it() bf(File list transfer time) is the number of seconds that the sender
1580 spent sending the file list to the receiver.
1581 it() bf(Total bytes sent) is the count of all the bytes that rsync sent
1582 from the client side to the server side.
1583 it() bf(Total bytes received) is the count of all non-message bytes that
1584 rsync received by the client side from the server side. "Non-message"
1585 bytes means that we don't count the bytes for a verbose message that the
1586 server sent to us, which makes the stats more consistent.
1589 dit(bf(-8, --8-bit-output)) This tells rsync to leave all high-bit characters
1590 unescaped in the output instead of trying to test them to see if they're
1591 valid in the current locale and escaping the invalid ones. All control
1592 characters (but never tabs) are always escaped, regardless of this option's
1595 The escape idiom that started in 2.6.7 is to output a literal backslash (\)
1596 and a hash (#), followed by exactly 3 octal digits. For example, a newline
1597 would output as "\#012". A literal backslash that is in a filename is not
1598 escaped unless it is followed by a hash and 3 digits (0-9).
1600 dit(bf(-h, --human-readable)) Output numbers in a more human-readable format.
1601 This makes big numbers output using larger units, with a K, M, or G suffix. If
1602 this option was specified once, these units are K (1000), M (1000*1000), and
1603 G (1000*1000*1000); if the option is repeated, the units are powers of 1024
1606 dit(bf(--partial)) By default, rsync will delete any partially
1607 transferred file if the transfer is interrupted. In some circumstances
1608 it is more desirable to keep partially transferred files. Using the
1609 bf(--partial) option tells rsync to keep the partial file which should
1610 make a subsequent transfer of the rest of the file much faster.
1612 dit(bf(--partial-dir=DIR)) A better way to keep partial files than the
1613 bf(--partial) option is to specify a em(DIR) that will be used to hold the
1614 partial data (instead of writing it out to the destination file).
1615 On the next transfer, rsync will use a file found in this
1616 dir as data to speed up the resumption of the transfer and then delete it
1617 after it has served its purpose.
1619 Note that if bf(--whole-file) is specified (or implied), any partial-dir
1620 file that is found for a file that is being updated will simply be removed
1622 rsync is sending files without using the incremental rsync algorithm).
1624 Rsync will create the em(DIR) if it is missing (just the last dir -- not
1625 the whole path). This makes it easy to use a relative path (such as
1626 "bf(--partial-dir=.rsync-partial)") to have rsync create the
1627 partial-directory in the destination file's directory when needed, and then
1628 remove it again when the partial file is deleted.
1630 If the partial-dir value is not an absolute path, rsync will add an exclude
1631 rule at the end of all your existing excludes. This will prevent the
1632 sending of any partial-dir files that may exist on the sending side, and
1633 will also prevent the untimely deletion of partial-dir items on the
1634 receiving side. An example: the above bf(--partial-dir) option would add
1635 the equivalent of "bf(-f '-p .rsync-partial/')" at the end of any other
1638 If you are supplying your own exclude rules, you may need to add your own
1639 exclude/hide/protect rule for the partial-dir because (1) the auto-added
1640 rule may be ineffective at the end of your other rules, or (2) you may wish
1641 to override rsync's exclude choice. For instance, if you want to make
1642 rsync clean-up any left-over partial-dirs that may be lying around, you
1643 should specify bf(--delete-after) and add a "risk" filter rule, e.g.
1644 bf(-f 'R .rsync-partial/'). (Avoid using bf(--delete-before) or
1645 bf(--delete-during) unless you don't need rsync to use any of the
1646 left-over partial-dir data during the current run.)
1648 IMPORTANT: the bf(--partial-dir) should not be writable by other users or it
1649 is a security risk. E.g. AVOID "/tmp".
1651 You can also set the partial-dir value the RSYNC_PARTIAL_DIR environment
1652 variable. Setting this in the environment does not force bf(--partial) to be
1653 enabled, but rather it affects where partial files go when bf(--partial) is
1654 specified. For instance, instead of using bf(--partial-dir=.rsync-tmp)
1655 along with bf(--progress), you could set RSYNC_PARTIAL_DIR=.rsync-tmp in your
1656 environment and then just use the bf(-P) option to turn on the use of the
1657 .rsync-tmp dir for partial transfers. The only times that the bf(--partial)
1658 option does not look for this environment value are (1) when bf(--inplace) was
1659 specified (since bf(--inplace) conflicts with bf(--partial-dir)), and (2) when
1660 bf(--delay-updates) was specified (see below).
1662 For the purposes of the daemon-config's "refuse options" setting,
1663 bf(--partial-dir) does em(not) imply bf(--partial). This is so that a
1664 refusal of the bf(--partial) option can be used to disallow the overwriting
1665 of destination files with a partial transfer, while still allowing the
1666 safer idiom provided by bf(--partial-dir).
1668 dit(bf(--delay-updates)) This option puts the temporary file from each
1669 updated file into a holding directory until the end of the
1670 transfer, at which time all the files are renamed into place in rapid
1671 succession. This attempts to make the updating of the files a little more
1672 atomic. By default the files are placed into a directory named ".~tmp~" in
1673 each file's destination directory, but if you've specified the
1674 bf(--partial-dir) option, that directory will be used instead. See the
1675 comments in the bf(--partial-dir) section for a discussion of how this
1676 ".~tmp~" dir will be excluded from the transfer, and what you can do if
1677 you want rsync to cleanup old ".~tmp~" dirs that might be lying around.
1678 Conflicts with bf(--inplace) and bf(--append).
1680 This option uses more memory on the receiving side (one bit per file
1681 transferred) and also requires enough free disk space on the receiving
1682 side to hold an additional copy of all the updated files. Note also that
1683 you should not use an absolute path to bf(--partial-dir) unless (1)
1685 chance of any of the files in the transfer having the same name (since all
1686 the updated files will be put into a single directory if the path is
1688 and (2) there are no mount points in the hierarchy (since the
1689 delayed updates will fail if they can't be renamed into place).
1691 See also the "atomic-rsync" perl script in the "support" subdir for an
1692 update algorithm that is even more atomic (it uses bf(--link-dest) and a
1693 parallel hierarchy of files).
1695 dit(bf(-m, --prune-empty-dirs)) This option tells the receiving rsync to get
1696 rid of empty directories from the file-list, including nested directories
1697 that have no non-directory children. This is useful for avoiding the
1698 creation of a bunch of useless directories when the sending rsync is
1699 recursively scanning a hierarchy of files using include/exclude/filter
1702 Because the file-list is actually being pruned, this option also affects
1703 what directories get deleted when a delete is active. However, keep in
1704 mind that excluded files and directories can prevent existing items from
1705 being deleted (because an exclude hides source files and protects
1708 You can prevent the pruning of certain empty directories from the file-list
1709 by using a global "protect" filter. For instance, this option would ensure
1710 that the directory "emptydir" was kept in the file-list:
1712 quote( --filter 'protect emptydir/')
1714 Here's an example that copies all .pdf files in a hierarchy, only creating
1715 the necessary destination directories to hold the .pdf files, and ensures
1716 that any superfluous files and directories in the destination are removed
1717 (note the hide filter of non-directories being used instead of an exclude):
1719 quote( rsync -avm --del --include='*.pdf' -f 'hide,! */' src/ dest)
1721 If you didn't want to remove superfluous destination files, the more
1722 time-honored options of "bf(--include='*/' --exclude='*')" would work fine
1723 in place of the hide-filter (if that is more natural to you).
1725 dit(bf(--progress)) This option tells rsync to print information
1726 showing the progress of the transfer. This gives a bored user
1728 Implies bf(--verbose) if it wasn't already specified.
1730 While rsync is transferring a regular file, it updates a progress line that
1733 verb( 782448 63% 110.64kB/s 0:00:04)
1735 In this example, the receiver has reconstructed 782448 bytes or 63% of the
1736 sender's file, which is being reconstructed at a rate of 110.64 kilobytes
1737 per second, and the transfer will finish in 4 seconds if the current rate
1738 is maintained until the end.
1740 These statistics can be misleading if the incremental transfer algorithm is
1741 in use. For example, if the sender's file consists of the basis file
1742 followed by additional data, the reported rate will probably drop
1743 dramatically when the receiver gets to the literal data, and the transfer
1744 will probably take much longer to finish than the receiver estimated as it
1745 was finishing the matched part of the file.
1747 When the file transfer finishes, rsync replaces the progress line with a
1748 summary line that looks like this:
1750 verb( 1238099 100% 146.38kB/s 0:00:08 (xfer#5, to-check=169/396))
1752 In this example, the file was 1238099 bytes long in total, the average rate
1753 of transfer for the whole file was 146.38 kilobytes per second over the 8
1754 seconds that it took to complete, it was the 5th transfer of a regular file
1755 during the current rsync session, and there are 169 more files for the
1756 receiver to check (to see if they are up-to-date or not) remaining out of
1757 the 396 total files in the file-list.
1759 dit(bf(-P)) The bf(-P) option is equivalent to bf(--partial) bf(--progress). Its
1760 purpose is to make it much easier to specify these two options for a long
1761 transfer that may be interrupted.
1763 dit(bf(--password-file)) This option allows you to provide a password in a
1764 file for accessing an rsync daemon. The file must not be world readable.
1765 It should contain just the password as a single line.
1767 When accessing an rsync daemon using a remote shell as the transport, this
1768 option only comes into effect after the remote shell finishes its
1769 authentication (i.e. if you have also specified a password in the daemon's
1772 dit(bf(--list-only)) This option will cause the source files to be listed
1773 instead of transferred. This option is inferred if there is a single source
1774 arg and no destination specified, so its main uses are: (1) to turn a copy
1775 command that includes a
1776 destination arg into a file-listing command, (2) to be able to specify more
1777 than one local source arg (note: be sure to include the destination), or
1778 (3) to avoid the automatically added "bf(-r --exclude='/*/*')" options that
1779 rsync usually uses as a compatibility kluge when generating a non-recursive
1780 listing. Caution: keep in mind that a source arg with a wild-card is expanded
1781 by the shell into multiple args, so it is never safe to try to list such an arg
1782 without using this option. For example:
1784 verb( rsync -av --list-only foo* dest/)
1786 dit(bf(--bwlimit=KBPS)) This option allows you to specify a maximum
1787 transfer rate in kilobytes per second. This option is most effective when
1788 using rsync with large files (several megabytes and up). Due to the nature
1789 of rsync transfers, blocks of data are sent, then if rsync determines the
1790 transfer was too fast, it will wait before sending the next data block. The
1791 result is an average transfer rate equaling the specified limit. A value
1792 of zero specifies no limit.
1794 dit(bf(--write-batch=FILE)) Record a file that can later be applied to
1795 another identical destination with bf(--read-batch). See the "BATCH MODE"
1796 section for details, and also the bf(--only-write-batch) option.
1798 dit(bf(--only-write-batch=FILE)) Works like bf(--write-batch), except that
1799 no updates are made on the destination system when creating the batch.
1800 This lets you transport the changes to the destination system via some
1801 other means and then apply the changes via bf(--read-batch).
1803 Note that you can feel free to write the batch directly to some portable
1804 media: if this media fills to capacity before the end of the transfer, you
1805 can just apply that partial transfer to the destination and repeat the
1806 whole process to get the rest of the changes (as long as you don't mind a
1807 partially updated destination system while the multi-update cycle is
1810 Also note that you only save bandwidth when pushing changes to a remote
1811 system because this allows the batched data to be diverted from the sender
1812 into the batch file without having to flow over the wire to the receiver
1813 (when pulling, the sender is remote, and thus can't write the batch).
1815 dit(bf(--read-batch=FILE)) Apply all of the changes stored in FILE, a
1816 file previously generated by bf(--write-batch).
1817 If em(FILE) is bf(-), the batch data will be read from standard input.
1818 See the "BATCH MODE" section for details.
1820 dit(bf(--protocol=NUM)) Force an older protocol version to be used. This
1821 is useful for creating a batch file that is compatible with an older
1822 version of rsync. For instance, if rsync 2.6.4 is being used with the
1823 bf(--write-batch) option, but rsync 2.6.3 is what will be used to run the
1824 bf(--read-batch) option, you should use "--protocol=28" when creating the
1825 batch file to force the older protocol version to be used in the batch
1826 file (assuming you can't upgrade the rsync on the reading system).
1828 dit(bf(-4, --ipv4) or bf(-6, --ipv6)) Tells rsync to prefer IPv4/IPv6
1829 when creating sockets. This only affects sockets that rsync has direct
1830 control over, such as the outgoing socket when directly contacting an
1831 rsync daemon. See also these options in the bf(--daemon) mode section.
1833 dit(bf(--checksum-seed=NUM)) Set the MD4 checksum seed to the integer
1834 NUM. This 4 byte checksum seed is included in each block and file
1835 MD4 checksum calculation. By default the checksum seed is generated
1836 by the server and defaults to the current code(time()). This option
1837 is used to set a specific checksum seed, which is useful for
1838 applications that want repeatable block and file checksums, or
1839 in the case where the user wants a more random checksum seed.
1840 Note that setting NUM to 0 causes rsync to use the default of code(time())
1844 manpagesection(DAEMON OPTIONS)
1846 The options allowed when starting an rsync daemon are as follows:
1849 dit(bf(--daemon)) This tells rsync that it is to run as a daemon. The
1850 daemon you start running may be accessed using an rsync client using
1851 the bf(host::module) or bf(rsync://host/module/) syntax.
1853 If standard input is a socket then rsync will assume that it is being
1854 run via inetd, otherwise it will detach from the current terminal and
1855 become a background daemon. The daemon will read the config file
1856 (rsyncd.conf) on each connect made by a client and respond to
1857 requests accordingly. See the bf(rsyncd.conf)(5) man page for more
1860 dit(bf(--address)) By default rsync will bind to the wildcard address when
1861 run as a daemon with the bf(--daemon) option. The bf(--address) option
1862 allows you to specify a specific IP address (or hostname) to bind to. This
1863 makes virtual hosting possible in conjunction with the bf(--config) option.
1864 See also the "address" global option in the rsyncd.conf manpage.
1866 dit(bf(--bwlimit=KBPS)) This option allows you to specify a maximum
1867 transfer rate in kilobytes per second for the data the daemon sends.
1868 The client can still specify a smaller bf(--bwlimit) value, but their
1869 requested value will be rounded down if they try to exceed it. See the
1870 client version of this option (above) for some extra details.
1872 dit(bf(--config=FILE)) This specifies an alternate config file than
1873 the default. This is only relevant when bf(--daemon) is specified.
1874 The default is /etc/rsyncd.conf unless the daemon is running over
1875 a remote shell program and the remote user is not the super-user; in that case
1876 the default is rsyncd.conf in the current directory (typically $HOME).
1878 dit(bf(--no-detach)) When running as a daemon, this option instructs
1879 rsync to not detach itself and become a background process. This
1880 option is required when running as a service on Cygwin, and may also
1881 be useful when rsync is supervised by a program such as
1882 bf(daemontools) or AIX's bf(System Resource Controller).
1883 bf(--no-detach) is also recommended when rsync is run under a
1884 debugger. This option has no effect if rsync is run from inetd or
1887 dit(bf(--port=PORT)) This specifies an alternate TCP port number for the
1888 daemon to listen on rather than the default of 873. See also the "port"
1889 global option in the rsyncd.conf manpage.
1891 dit(bf(--log-file=FILE)) This option tells the rsync daemon to use the
1892 given log-file name instead of using the "log file" setting in the config
1895 dit(bf(--log-file-format=FORMAT)) This option tells the rsync daemon to use the
1896 given FORMAT string instead of using the "log format" setting in the config
1897 file. It also enables "transfer logging" unless the string is empty, in which
1898 case transfer logging is turned off.
1900 dit(bf(--sockopts)) This overrides the bf(socket options) setting in the
1901 rsyncd.conf file and has the same syntax.
1903 dit(bf(-v, --verbose)) This option increases the amount of information the
1904 daemon logs during its startup phase. After the client connects, the
1905 daemon's verbosity level will be controlled by the options that the client
1906 used and the "max verbosity" setting in the module's config section.
1908 dit(bf(-4, --ipv4) or bf(-6, --ipv6)) Tells rsync to prefer IPv4/IPv6
1909 when creating the incoming sockets that the rsync daemon will use to
1910 listen for connections. One of these options may be required in older
1911 versions of Linux to work around an IPv6 bug in the kernel (if you see
1912 an "address already in use" error when nothing else is using the port,
1913 try specifying bf(--ipv6) or bf(--ipv4) when starting the daemon).
1915 dit(bf(-h, --help)) When specified after bf(--daemon), print a short help
1916 page describing the options available for starting an rsync daemon.
1919 manpagesection(FILTER RULES)
1921 The filter rules allow for flexible selection of which files to transfer
1922 (include) and which files to skip (exclude). The rules either directly
1923 specify include/exclude patterns or they specify a way to acquire more
1924 include/exclude patterns (e.g. to read them from a file).
1926 As the list of files/directories to transfer is built, rsync checks each
1927 name to be transferred against the list of include/exclude patterns in
1928 turn, and the first matching pattern is acted on: if it is an exclude
1929 pattern, then that file is skipped; if it is an include pattern then that
1930 filename is not skipped; if no matching pattern is found, then the
1931 filename is not skipped.
1933 Rsync builds an ordered list of filter rules as specified on the
1934 command-line. Filter rules have the following syntax:
1937 tt(RULE [PATTERN_OR_FILENAME])nl()
1938 tt(RULE,MODIFIERS [PATTERN_OR_FILENAME])nl()
1941 You have your choice of using either short or long RULE names, as described
1942 below. If you use a short-named rule, the ',' separating the RULE from the
1943 MODIFIERS is optional. The PATTERN or FILENAME that follows (when present)
1944 must come after either a single space or an underscore (_).
1945 Here are the available rule prefixes:
1948 bf(exclude, -) specifies an exclude pattern. nl()
1949 bf(include, +) specifies an include pattern. nl()
1950 bf(merge, .) specifies a merge-file to read for more rules. nl()
1951 bf(dir-merge, :) specifies a per-directory merge-file. nl()
1952 bf(hide, H) specifies a pattern for hiding files from the transfer. nl()
1953 bf(show, S) files that match the pattern are not hidden. nl()
1954 bf(protect, P) specifies a pattern for protecting files from deletion. nl()
1955 bf(risk, R) files that match the pattern are not protected. nl()
1956 bf(clear, !) clears the current include/exclude list (takes no arg) nl()
1959 When rules are being read from a file, empty lines are ignored, as are
1960 comment lines that start with a "#".
1962 Note that the bf(--include)/bf(--exclude) command-line options do not allow the
1963 full range of rule parsing as described above -- they only allow the
1964 specification of include/exclude patterns plus a "!" token to clear the
1965 list (and the normal comment parsing when rules are read from a file).
1967 does not begin with "- " (dash, space) or "+ " (plus, space), then the
1968 rule will be interpreted as if "+ " (for an include option) or "- " (for
1969 an exclude option) were prefixed to the string. A bf(--filter) option, on
1970 the other hand, must always contain either a short or long rule name at the
1973 Note also that the bf(--filter), bf(--include), and bf(--exclude) options take one
1974 rule/pattern each. To add multiple ones, you can repeat the options on
1975 the command-line, use the merge-file syntax of the bf(--filter) option, or
1976 the bf(--include-from)/bf(--exclude-from) options.
1978 manpagesection(INCLUDE/EXCLUDE PATTERN RULES)
1980 You can include and exclude files by specifying patterns using the "+",
1981 "-", etc. filter rules (as introduced in the FILTER RULES section above).
1982 The include/exclude rules each specify a pattern that is matched against
1983 the names of the files that are going to be transferred. These patterns
1984 can take several forms:
1987 it() if the pattern starts with a / then it is anchored to a
1988 particular spot in the hierarchy of files, otherwise it is matched
1989 against the end of the pathname. This is similar to a leading ^ in
1990 regular expressions.
1991 Thus "/foo" would match a file named "foo" at either the "root of the
1992 transfer" (for a global rule) or in the merge-file's directory (for a
1993 per-directory rule).
1994 An unqualified "foo" would match any file or directory named "foo"
1995 anywhere in the tree because the algorithm is applied recursively from
1997 top down; it behaves as if each path component gets a turn at being the
1998 end of the file name. Even the unanchored "sub/foo" would match at
1999 any point in the hierarchy where a "foo" was found within a directory
2000 named "sub". See the section on ANCHORING INCLUDE/EXCLUDE PATTERNS for
2001 a full discussion of how to specify a pattern that matches at the root
2003 it() if the pattern ends with a / then it will only match a
2004 directory, not a file, link, or device.
2005 it() rsync chooses between doing a simple string match and wildcard
2006 matching by checking if the pattern contains one of these three wildcard
2007 characters: '*', '?', and '[' .
2008 it() a '*' matches any non-empty path component (it stops at slashes).
2009 it() use '**' to match anything, including slashes.
2010 it() a '?' matches any character except a slash (/).
2011 it() a '[' introduces a character class, such as [a-z] or [[:alpha:]].
2012 it() in a wildcard pattern, a backslash can be used to escape a wildcard
2013 character, but it is matched literally when no wildcards are present.
2014 it() if the pattern contains a / (not counting a trailing /) or a "**",
2015 then it is matched against the full pathname, including any leading
2016 directories. If the pattern doesn't contain a / or a "**", then it is
2017 matched only against the final component of the filename.
2018 (Remember that the algorithm is applied recursively so "full filename"
2019 can actually be any portion of a path from the starting directory on
2021 it() a trailing "dir_name/***" will match both the directory (as if
2022 "dir_name/" had been specified) and all the files in the directory
2023 (as if "dir_name/**" had been specified). This behavior was added in
2027 Note that, when using the bf(--recursive) (bf(-r)) option (which is implied by
2028 bf(-a)), every subcomponent of every path is visited from the top down, so
2029 include/exclude patterns get applied recursively to each subcomponent's
2030 full name (e.g. to include "/foo/bar/baz" the subcomponents "/foo" and
2031 "/foo/bar" must not be excluded).
2032 The exclude patterns actually short-circuit the directory traversal stage
2033 when rsync finds the files to send. If a pattern excludes a particular
2034 parent directory, it can render a deeper include pattern ineffectual
2035 because rsync did not descend through that excluded section of the
2036 hierarchy. This is particularly important when using a trailing '*' rule.
2037 For instance, this won't work:
2040 tt(+ /some/path/this-file-will-not-be-found)nl()
2041 tt(+ /file-is-included)nl()
2045 This fails because the parent directory "some" is excluded by the '*'
2046 rule, so rsync never visits any of the files in the "some" or "some/path"
2047 directories. One solution is to ask for all directories in the hierarchy
2048 to be included by using a single rule: "+ */" (put it somewhere before the
2049 "- *" rule), and perhaps use the bf(--prune-empty-dirs) option. Another
2050 solution is to add specific include rules for all
2051 the parent dirs that need to be visited. For instance, this set of rules
2056 tt(+ /some/path/)nl()
2057 tt(+ /some/path/this-file-is-found)nl()
2058 tt(+ /file-also-included)nl()
2062 Here are some examples of exclude/include matching:
2065 it() "- *.o" would exclude all filenames matching *.o
2066 it() "- /foo" would exclude a file (or directory) named foo in the
2067 transfer-root directory
2068 it() "- foo/" would exclude any directory named foo
2069 it() "- /foo/*/bar" would exclude any file named bar which is at two
2070 levels below a directory named foo in the transfer-root directory
2071 it() "- /foo/**/bar" would exclude any file named bar two
2072 or more levels below a directory named foo in the transfer-root directory
2073 it() The combination of "+ */", "+ *.c", and "- *" would include all
2074 directories and C source files but nothing else (see also the
2075 bf(--prune-empty-dirs) option)
2076 it() The combination of "+ foo/", "+ foo/bar.c", and "- *" would include
2077 only the foo directory and foo/bar.c (the foo directory must be
2078 explicitly included or it would be excluded by the "*")
2081 manpagesection(MERGE-FILE FILTER RULES)
2083 You can merge whole files into your filter rules by specifying either a
2084 merge (.) or a dir-merge (:) filter rule (as introduced in the FILTER RULES
2087 There are two kinds of merged files -- single-instance ('.') and
2088 per-directory (':'). A single-instance merge file is read one time, and
2089 its rules are incorporated into the filter list in the place of the "."
2090 rule. For per-directory merge files, rsync will scan every directory that
2091 it traverses for the named file, merging its contents when the file exists
2092 into the current list of inherited rules. These per-directory rule files
2093 must be created on the sending side because it is the sending side that is
2094 being scanned for the available files to transfer. These rule files may
2095 also need to be transferred to the receiving side if you want them to
2096 affect what files don't get deleted (see PER-DIRECTORY RULES AND DELETE
2102 tt(merge /etc/rsync/default.rules)nl()
2103 tt(. /etc/rsync/default.rules)nl()
2104 tt(dir-merge .per-dir-filter)nl()
2105 tt(dir-merge,n- .non-inherited-per-dir-excludes)nl()
2106 tt(:n- .non-inherited-per-dir-excludes)nl()
2109 The following modifiers are accepted after a merge or dir-merge rule:
2112 it() A bf(-) specifies that the file should consist of only exclude
2113 patterns, with no other rule-parsing except for in-file comments.
2114 it() A bf(+) specifies that the file should consist of only include
2115 patterns, with no other rule-parsing except for in-file comments.
2116 it() A bf(C) is a way to specify that the file should be read in a
2117 CVS-compatible manner. This turns on 'n', 'w', and '-', but also
2118 allows the list-clearing token (!) to be specified. If no filename is
2119 provided, ".cvsignore" is assumed.
2120 it() A bf(e) will exclude the merge-file name from the transfer; e.g.
2121 "dir-merge,e .rules" is like "dir-merge .rules" and "- .rules".
2122 it() An bf(n) specifies that the rules are not inherited by subdirectories.
2123 it() A bf(w) specifies that the rules are word-split on whitespace instead
2124 of the normal line-splitting. This also turns off comments. Note: the
2125 space that separates the prefix from the rule is treated specially, so
2126 "- foo + bar" is parsed as two rules (assuming that prefix-parsing wasn't
2128 it() You may also specify any of the modifiers for the "+" or "-" rules
2129 (below) in order to have the rules that are read in from the file
2130 default to having that modifier set. For instance, "merge,-/ .excl" would
2131 treat the contents of .excl as absolute-path excludes,
2132 while "dir-merge,s .filt" and ":sC" would each make all their
2133 per-directory rules apply only on the sending side.
2136 The following modifiers are accepted after a "+" or "-":
2139 it() A bf(/) specifies that the include/exclude rule should be matched
2140 against the absolute pathname of the current item. For example,
2141 "-/ /etc/passwd" would exclude the passwd file any time the transfer
2142 was sending files from the "/etc" directory, and "-/ subdir/foo"
2143 would always exclude "foo" when it is in a dir named "subdir", even
2144 if "foo" is at the root of the current transfer.
2145 it() A bf(!) specifies that the include/exclude should take effect if
2146 the pattern fails to match. For instance, "-! */" would exclude all
2148 it() A bf(C) is used to indicate that all the global CVS-exclude rules
2149 should be inserted as excludes in place of the "-C". No arg should
2151 it() An bf(s) is used to indicate that the rule applies to the sending
2152 side. When a rule affects the sending side, it prevents files from
2153 being transferred. The default is for a rule to affect both sides
2154 unless bf(--delete-excluded) was specified, in which case default rules
2155 become sender-side only. See also the hide (H) and show (S) rules,
2156 which are an alternate way to specify sending-side includes/excludes.
2157 it() An bf(r) is used to indicate that the rule applies to the receiving
2158 side. When a rule affects the receiving side, it prevents files from
2159 being deleted. See the bf(s) modifier for more info. See also the
2160 protect (P) and risk (R) rules, which are an alternate way to
2161 specify receiver-side includes/excludes.
2162 it() A bf(p) indicates that a rule is perishable, meaning that it is
2163 ignored in directories that are being deleted. For instance, the bf(-C)
2164 option's default rules that exclude things like "CVS" and "*.o" are
2165 marked as perishable, and will not prevent a directory that was removed
2166 on the source from being deleted on the destination.
2169 Per-directory rules are inherited in all subdirectories of the directory
2170 where the merge-file was found unless the 'n' modifier was used. Each
2171 subdirectory's rules are prefixed to the inherited per-directory rules
2172 from its parents, which gives the newest rules a higher priority than the
2173 inherited rules. The entire set of dir-merge rules are grouped together in
2174 the spot where the merge-file was specified, so it is possible to override
2175 dir-merge rules via a rule that got specified earlier in the list of global
2176 rules. When the list-clearing rule ("!") is read from a per-directory
2177 file, it only clears the inherited rules for the current merge file.
2179 Another way to prevent a single rule from a dir-merge file from being inherited is to
2180 anchor it with a leading slash. Anchored rules in a per-directory
2181 merge-file are relative to the merge-file's directory, so a pattern "/foo"
2182 would only match the file "foo" in the directory where the dir-merge filter
2185 Here's an example filter file which you'd specify via bf(--filter=". file":)
2188 tt(merge /home/user/.global-filter)nl()
2190 tt(dir-merge .rules)nl()
2195 This will merge the contents of the /home/user/.global-filter file at the
2196 start of the list and also turns the ".rules" filename into a per-directory
2197 filter file. All rules read in prior to the start of the directory scan
2198 follow the global anchoring rules (i.e. a leading slash matches at the root
2201 If a per-directory merge-file is specified with a path that is a parent
2202 directory of the first transfer directory, rsync will scan all the parent
2203 dirs from that starting point to the transfer directory for the indicated
2204 per-directory file. For instance, here is a common filter (see bf(-F)):
2206 quote(tt(--filter=': /.rsync-filter'))
2208 That rule tells rsync to scan for the file .rsync-filter in all
2209 directories from the root down through the parent directory of the
2210 transfer prior to the start of the normal directory scan of the file in
2211 the directories that are sent as a part of the transfer. (Note: for an
2212 rsync daemon, the root is always the same as the module's "path".)
2214 Some examples of this pre-scanning for per-directory files:
2217 tt(rsync -avF /src/path/ /dest/dir)nl()
2218 tt(rsync -av --filter=': ../../.rsync-filter' /src/path/ /dest/dir)nl()
2219 tt(rsync -av --filter=': .rsync-filter' /src/path/ /dest/dir)nl()
2222 The first two commands above will look for ".rsync-filter" in "/" and
2223 "/src" before the normal scan begins looking for the file in "/src/path"
2224 and its subdirectories. The last command avoids the parent-dir scan
2225 and only looks for the ".rsync-filter" files in each directory that is
2226 a part of the transfer.
2228 If you want to include the contents of a ".cvsignore" in your patterns,
2229 you should use the rule ":C", which creates a dir-merge of the .cvsignore
2230 file, but parsed in a CVS-compatible manner. You can
2231 use this to affect where the bf(--cvs-exclude) (bf(-C)) option's inclusion of the
2232 per-directory .cvsignore file gets placed into your rules by putting the
2233 ":C" wherever you like in your filter rules. Without this, rsync would
2234 add the dir-merge rule for the .cvsignore file at the end of all your other
2235 rules (giving it a lower priority than your command-line rules). For
2239 tt(cat <<EOT | rsync -avC --filter='. -' a/ b)nl()
2244 tt(rsync -avC --include=foo.o -f :C --exclude='*.old' a/ b)nl()
2247 Both of the above rsync commands are identical. Each one will merge all
2248 the per-directory .cvsignore rules in the middle of the list rather than
2249 at the end. This allows their dir-specific rules to supersede the rules
2250 that follow the :C instead of being subservient to all your rules. To
2251 affect the other CVS exclude rules (i.e. the default list of exclusions,
2252 the contents of $HOME/.cvsignore, and the value of $CVSIGNORE) you should
2253 omit the bf(-C) command-line option and instead insert a "-C" rule into
2254 your filter rules; e.g. "bf(--filter=-C)".
2256 manpagesection(LIST-CLEARING FILTER RULE)
2258 You can clear the current include/exclude list by using the "!" filter
2259 rule (as introduced in the FILTER RULES section above). The "current"
2260 list is either the global list of rules (if the rule is encountered while
2261 parsing the filter options) or a set of per-directory rules (which are
2262 inherited in their own sub-list, so a subdirectory can use this to clear
2263 out the parent's rules).
2265 manpagesection(ANCHORING INCLUDE/EXCLUDE PATTERNS)
2267 As mentioned earlier, global include/exclude patterns are anchored at the
2268 "root of the transfer" (as opposed to per-directory patterns, which are
2269 anchored at the merge-file's directory). If you think of the transfer as
2270 a subtree of names that are being sent from sender to receiver, the
2271 transfer-root is where the tree starts to be duplicated in the destination
2272 directory. This root governs where patterns that start with a / match.
2274 Because the matching is relative to the transfer-root, changing the
2275 trailing slash on a source path or changing your use of the bf(--relative)
2276 option affects the path you need to use in your matching (in addition to
2277 changing how much of the file tree is duplicated on the destination
2278 host). The following examples demonstrate this.
2280 Let's say that we want to match two source files, one with an absolute
2281 path of "/home/me/foo/bar", and one with a path of "/home/you/bar/baz".
2282 Here is how the various command choices differ for a 2-source transfer:
2285 Example cmd: rsync -a /home/me /home/you /dest nl()
2286 +/- pattern: /me/foo/bar nl()
2287 +/- pattern: /you/bar/baz nl()
2288 Target file: /dest/me/foo/bar nl()
2289 Target file: /dest/you/bar/baz nl()
2293 Example cmd: rsync -a /home/me/ /home/you/ /dest nl()
2294 +/- pattern: /foo/bar (note missing "me") nl()
2295 +/- pattern: /bar/baz (note missing "you") nl()
2296 Target file: /dest/foo/bar nl()
2297 Target file: /dest/bar/baz nl()
2301 Example cmd: rsync -a --relative /home/me/ /home/you /dest nl()
2302 +/- pattern: /home/me/foo/bar (note full path) nl()
2303 +/- pattern: /home/you/bar/baz (ditto) nl()
2304 Target file: /dest/home/me/foo/bar nl()
2305 Target file: /dest/home/you/bar/baz nl()
2309 Example cmd: cd /home; rsync -a --relative me/foo you/ /dest nl()
2310 +/- pattern: /me/foo/bar (starts at specified path) nl()
2311 +/- pattern: /you/bar/baz (ditto) nl()
2312 Target file: /dest/me/foo/bar nl()
2313 Target file: /dest/you/bar/baz nl()
2316 The easiest way to see what name you should filter is to just
2317 look at the output when using bf(--verbose) and put a / in front of the name
2318 (use the bf(--dry-run) option if you're not yet ready to copy any files).
2320 manpagesection(PER-DIRECTORY RULES AND DELETE)
2322 Without a delete option, per-directory rules are only relevant on the
2323 sending side, so you can feel free to exclude the merge files themselves
2324 without affecting the transfer. To make this easy, the 'e' modifier adds
2325 this exclude for you, as seen in these two equivalent commands:
2328 tt(rsync -av --filter=': .excl' --exclude=.excl host:src/dir /dest)nl()
2329 tt(rsync -av --filter=':e .excl' host:src/dir /dest)nl()
2332 However, if you want to do a delete on the receiving side AND you want some
2333 files to be excluded from being deleted, you'll need to be sure that the
2334 receiving side knows what files to exclude. The easiest way is to include
2335 the per-directory merge files in the transfer and use bf(--delete-after),
2336 because this ensures that the receiving side gets all the same exclude
2337 rules as the sending side before it tries to delete anything:
2339 quote(tt(rsync -avF --delete-after host:src/dir /dest))
2341 However, if the merge files are not a part of the transfer, you'll need to
2342 either specify some global exclude rules (i.e. specified on the command
2343 line), or you'll need to maintain your own per-directory merge files on
2344 the receiving side. An example of the first is this (assume that the
2345 remote .rules files exclude themselves):
2347 verb(rsync -av --filter=': .rules' --filter='. /my/extra.rules'
2348 --delete host:src/dir /dest)
2350 In the above example the extra.rules file can affect both sides of the
2351 transfer, but (on the sending side) the rules are subservient to the rules
2352 merged from the .rules files because they were specified after the
2353 per-directory merge rule.
2355 In one final example, the remote side is excluding the .rsync-filter
2356 files from the transfer, but we want to use our own .rsync-filter files
2357 to control what gets deleted on the receiving side. To do this we must
2358 specifically exclude the per-directory merge files (so that they don't get
2359 deleted) and then put rules into the local files to control what else
2360 should not get deleted. Like one of these commands:
2362 verb( rsync -av --filter=':e /.rsync-filter' --delete \
2364 rsync -avFF --delete host:src/dir /dest)
2366 manpagesection(BATCH MODE)
2368 Batch mode can be used to apply the same set of updates to many
2369 identical systems. Suppose one has a tree which is replicated on a
2370 number of hosts. Now suppose some changes have been made to this
2371 source tree and those changes need to be propagated to the other
2372 hosts. In order to do this using batch mode, rsync is run with the
2373 write-batch option to apply the changes made to the source tree to one
2374 of the destination trees. The write-batch option causes the rsync
2375 client to store in a "batch file" all the information needed to repeat
2376 this operation against other, identical destination trees.
2378 To apply the recorded changes to another destination tree, run rsync
2379 with the read-batch option, specifying the name of the same batch
2380 file, and the destination tree. Rsync updates the destination tree
2381 using the information stored in the batch file.
2383 For convenience, one additional file is creating when the write-batch
2384 option is used. This file's name is created by appending
2385 ".sh" to the batch filename. The .sh file contains
2386 a command-line suitable for updating a destination tree using that
2387 batch file. It can be executed using a Bourne (or Bourne-like) shell,
2389 passing in an alternate destination tree pathname which is then used
2390 instead of the original path. This is useful when the destination tree
2391 path differs from the original destination tree path.
2393 Generating the batch file once saves having to perform the file
2394 status, checksum, and data block generation more than once when
2395 updating multiple destination trees. Multicast transport protocols can
2396 be used to transfer the batch update files in parallel to many hosts
2397 at once, instead of sending the same data to every host individually.
2402 tt($ rsync --write-batch=foo -a host:/source/dir/ /adest/dir/)nl()
2403 tt($ scp foo* remote:)nl()
2404 tt($ ssh remote ./foo.sh /bdest/dir/)nl()
2408 tt($ rsync --write-batch=foo -a /source/dir/ /adest/dir/)nl()
2409 tt($ ssh remote rsync --read-batch=- -a /bdest/dir/ <foo)nl()
2412 In these examples, rsync is used to update /adest/dir/ from /source/dir/
2413 and the information to repeat this operation is stored in "foo" and
2414 "foo.sh". The host "remote" is then updated with the batched data going
2415 into the directory /bdest/dir. The differences between the two examples
2416 reveals some of the flexibility you have in how you deal with batches:
2419 it() The first example shows that the initial copy doesn't have to be
2420 local -- you can push or pull data to/from a remote host using either the
2421 remote-shell syntax or rsync daemon syntax, as desired.
2422 it() The first example uses the created "foo.sh" file to get the right
2423 rsync options when running the read-batch command on the remote host.
2424 it() The second example reads the batch data via standard input so that
2425 the batch file doesn't need to be copied to the remote machine first.
2426 This example avoids the foo.sh script because it needed to use a modified
2427 bf(--read-batch) option, but you could edit the script file if you wished to
2428 make use of it (just be sure that no other option is trying to use
2429 standard input, such as the "bf(--exclude-from=-)" option).
2434 The read-batch option expects the destination tree that it is updating
2435 to be identical to the destination tree that was used to create the
2436 batch update fileset. When a difference between the destination trees
2437 is encountered the update might be discarded with a warning (if the file
2438 appears to be up-to-date already) or the file-update may be attempted
2439 and then, if the file fails to verify, the update discarded with an
2440 error. This means that it should be safe to re-run a read-batch operation
2441 if the command got interrupted. If you wish to force the batched-update to
2442 always be attempted regardless of the file's size and date, use the bf(-I)
2443 option (when reading the batch).
2444 If an error occurs, the destination tree will probably be in a
2445 partially updated state. In that case, rsync can
2446 be used in its regular (non-batch) mode of operation to fix up the
2449 The rsync version used on all destinations must be at least as new as the
2450 one used to generate the batch file. Rsync will die with an error if the
2451 protocol version in the batch file is too new for the batch-reading rsync
2452 to handle. See also the bf(--protocol) option for a way to have the
2453 creating rsync generate a batch file that an older rsync can understand.
2454 (Note that batch files changed format in version 2.6.3, so mixing versions
2455 older than that with newer versions will not work.)
2457 When reading a batch file, rsync will force the value of certain options
2458 to match the data in the batch file if you didn't set them to the same
2459 as the batch-writing command. Other options can (and should) be changed.
2460 For instance bf(--write-batch) changes to bf(--read-batch),
2461 bf(--files-from) is dropped, and the
2462 bf(--filter)/bf(--include)/bf(--exclude) options are not needed unless
2463 one of the bf(--delete) options is specified.
2465 The code that creates the BATCH.sh file transforms any filter/include/exclude
2466 options into a single list that is appended as a "here" document to the
2467 shell script file. An advanced user can use this to modify the exclude
2468 list if a change in what gets deleted by bf(--delete) is desired. A normal
2469 user can ignore this detail and just use the shell script as an easy way
2470 to run the appropriate bf(--read-batch) command for the batched data.
2472 The original batch mode in rsync was based on "rsync+", but the latest
2473 version uses a new implementation.
2475 manpagesection(SYMBOLIC LINKS)
2477 Three basic behaviors are possible when rsync encounters a symbolic
2478 link in the source directory.
2480 By default, symbolic links are not transferred at all. A message
2481 "skipping non-regular" file is emitted for any symlinks that exist.
2483 If bf(--links) is specified, then symlinks are recreated with the same
2484 target on the destination. Note that bf(--archive) implies
2487 If bf(--copy-links) is specified, then symlinks are "collapsed" by
2488 copying their referent, rather than the symlink.
2490 rsync also distinguishes "safe" and "unsafe" symbolic links. An
2491 example where this might be used is a web site mirror that wishes
2492 ensure the rsync module they copy does not include symbolic links to
2493 bf(/etc/passwd) in the public section of the site. Using
2494 bf(--copy-unsafe-links) will cause any links to be copied as the file
2495 they point to on the destination. Using bf(--safe-links) will cause
2496 unsafe links to be omitted altogether. (Note that you must specify
2497 bf(--links) for bf(--safe-links) to have any effect.)
2499 Symbolic links are considered unsafe if they are absolute symlinks
2500 (start with bf(/)), empty, or if they contain enough ".."
2501 components to ascend from the directory being copied.
2503 Here's a summary of how the symlink options are interpreted. The list is
2504 in order of precedence, so if your combination of options isn't mentioned,
2505 use the first line that is a complete subset of your options:
2507 dit(bf(--copy-links)) Turn all symlinks into normal files (leaving no
2508 symlinks for any other options to affect).
2510 dit(bf(--links --copy-unsafe-links)) Turn all unsafe symlinks into files
2511 and duplicate all safe symlinks.
2513 dit(bf(--copy-unsafe-links)) Turn all unsafe symlinks into files, noisily
2514 skip all safe symlinks.
2516 dit(bf(--links --safe-links)) Duplicate safe symlinks and skip unsafe
2519 dit(bf(--links)) Duplicate all symlinks.
2521 manpagediagnostics()
2523 rsync occasionally produces error messages that may seem a little
2524 cryptic. The one that seems to cause the most confusion is "protocol
2525 version mismatch -- is your shell clean?".
2527 This message is usually caused by your startup scripts or remote shell
2528 facility producing unwanted garbage on the stream that rsync is using
2529 for its transport. The way to diagnose this problem is to run your
2530 remote shell like this:
2532 quote(tt(ssh remotehost /bin/true > out.dat))
2534 then look at out.dat. If everything is working correctly then out.dat
2535 should be a zero length file. If you are getting the above error from
2536 rsync then you will probably find that out.dat contains some text or
2537 data. Look at the contents and try to work out what is producing
2538 it. The most common cause is incorrectly configured shell startup
2539 scripts (such as .cshrc or .profile) that contain output statements
2540 for non-interactive logins.
2542 If you are having trouble debugging filter patterns, then
2543 try specifying the bf(-vv) option. At this level of verbosity rsync will
2544 show why each individual file is included or excluded.
2546 manpagesection(EXIT VALUES)
2550 dit(bf(1)) Syntax or usage error
2551 dit(bf(2)) Protocol incompatibility
2552 dit(bf(3)) Errors selecting input/output files, dirs
2553 dit(bf(4)) Requested action not supported: an attempt
2554 was made to manipulate 64-bit files on a platform that cannot support
2555 them; or an option was specified that is supported by the client and
2557 dit(bf(5)) Error starting client-server protocol
2558 dit(bf(6)) Daemon unable to append to log-file
2559 dit(bf(10)) Error in socket I/O
2560 dit(bf(11)) Error in file I/O
2561 dit(bf(12)) Error in rsync protocol data stream
2562 dit(bf(13)) Errors with program diagnostics
2563 dit(bf(14)) Error in IPC code
2564 dit(bf(20)) Received SIGUSR1 or SIGINT
2565 dit(bf(21)) Some error returned by code(waitpid())
2566 dit(bf(22)) Error allocating core memory buffers
2567 dit(bf(23)) Partial transfer due to error
2568 dit(bf(24)) Partial transfer due to vanished source files
2569 dit(bf(25)) The --max-delete limit stopped deletions
2570 dit(bf(30)) Timeout in data send/receive
2573 manpagesection(ENVIRONMENT VARIABLES)
2576 dit(bf(CVSIGNORE)) The CVSIGNORE environment variable supplements any
2577 ignore patterns in .cvsignore files. See the bf(--cvs-exclude) option for
2579 dit(bf(RSYNC_RSH)) The RSYNC_RSH environment variable allows you to
2580 override the default shell used as the transport for rsync. Command line
2581 options are permitted after the command name, just as in the bf(-e) option.
2582 dit(bf(RSYNC_PROXY)) The RSYNC_PROXY environment variable allows you to
2583 redirect your rsync client to use a web proxy when connecting to a
2584 rsync daemon. You should set RSYNC_PROXY to a hostname:port pair.
2585 dit(bf(RSYNC_PASSWORD)) Setting RSYNC_PASSWORD to the required
2586 password allows you to run authenticated rsync connections to an rsync
2587 daemon without user intervention. Note that this does not supply a
2588 password to a shell transport such as ssh.
2589 dit(bf(USER) or bf(LOGNAME)) The USER or LOGNAME environment variables
2590 are used to determine the default username sent to an rsync daemon.
2591 If neither is set, the username defaults to "nobody".
2592 dit(bf(HOME)) The HOME environment variable is used to find the user's
2593 default .cvsignore file.
2598 /etc/rsyncd.conf or rsyncd.conf
2606 times are transferred as *nix time_t values
2608 When transferring to FAT filesystems rsync may re-sync
2610 See the comments on the bf(--modify-window) option.
2612 file permissions, devices, etc. are transferred as native numerical
2615 see also the comments on the bf(--delete) option
2617 Please report bugs! See the website at
2618 url(http://rsync.samba.org/)(http://rsync.samba.org/)
2620 manpagesection(VERSION)
2622 This man page is current for version 2.6.9 of rsync.
2624 manpagesection(INTERNAL OPTIONS)
2626 The options bf(--server) and bf(--sender) are used internally by rsync,
2627 and should never be typed by a user under normal circumstances. Some
2628 awareness of these options may be needed in certain scenarios, such as
2629 when setting up a login that can only run an rsync command. For instance,
2630 the support directory of the rsync distribution has an example script
2631 named rrsync (for restricted rsync) that can be used with a restricted
2634 manpagesection(CREDITS)
2636 rsync is distributed under the GNU public license. See the file
2637 COPYING for details.
2639 A WEB site is available at
2640 url(http://rsync.samba.org/)(http://rsync.samba.org/). The site
2641 includes an FAQ-O-Matic which may cover questions unanswered by this
2644 The primary ftp site for rsync is
2645 url(ftp://rsync.samba.org/pub/rsync)(ftp://rsync.samba.org/pub/rsync).
2647 We would be delighted to hear from you if you like this program.
2649 This program uses the excellent zlib compression library written by
2650 Jean-loup Gailly and Mark Adler.
2652 manpagesection(THANKS)
2654 Thanks to Richard Brent, Brendan Mackay, Bill Waite, Stephen Rothwell
2655 and David Bell for helpful suggestions, patches and testing of rsync.
2656 I've probably missed some people, my apologies if I have.
2658 Especial thanks also to: David Dykstra, Jos Backus, Sebastian Krahmer,
2659 Martin Pool, Wayne Davison, J.W. Schultz.
2663 rsync was originally written by Andrew Tridgell and Paul Mackerras.
2664 Many people have later contributed to it.
2666 Mailing lists for support and development are available at
2667 url(http://lists.samba.org)(lists.samba.org)