]> git.wh0rd.org - dump.git/blob - dump/dump.8.in
QFA fixes.
[dump.git] / dump / dump.8.in
1 .\" Copyright (c) 1980, 1991, 1993
2 .\" Regents of the University of California.
3 .\" All rights reserved.
4 .\"
5 .\" Redistribution and use in source and binary forms, with or without
6 .\" modification, are permitted provided that the following conditions
7 .\" are met:
8 .\" 1. Redistributions of source code must retain the above copyright
9 .\" notice, this list of conditions and the following disclaimer.
10 .\" 2. Redistributions in binary form must reproduce the above copyright
11 .\" notice, this list of conditions and the following disclaimer in the
12 .\" documentation and/or other materials provided with the distribution.
13 .\" 3. All advertising materials mentioning features or use of this software
14 .\" must display the following acknowledgement:
15 .\" This product includes software developed by the University of
16 .\" California, Berkeley and its contributors.
17 .\" 4. Neither the name of the University nor the names of its contributors
18 .\" may be used to endorse or promote products derived from this software
19 .\" without specific prior written permission.
20 .\"
21 .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31 .\" SUCH DAMAGE.
32 .\"
33 .\" $Id: dump.8.in,v 1.29 2001/07/18 13:12:33 stelian Exp $
34 .\"
35 .Dd __DATE__
36 .Dt DUMP 8
37 .Os "dump __VERSION__"
38 .Sh NAME
39 .Nm dump
40 .Nd ext2 filesystem backup
41 .Sh SYNOPSIS
42 .Nm dump
43 .Op Fl 0123456789ackMnSu
44 .Op Fl B Ar records
45 .Op Fl b Ar blocksize
46 .Op Fl d Ar density
47 .Op Fl e Ar inode number
48 .Op Fl f Ar file
49 .Op Fl F Ar script
50 .Op Fl h Ar level
51 .Op Fl L Ar label
52 .Op Fl Q Ar file
53 .Op Fl s Ar feet
54 .Op Fl T Ar date
55 .Op Fl z Ar compression level
56 .Ar files-to-dump
57 .Nm dump
58 .Op Fl W Li \&| Fl w
59 .Pp
60 .in
61 (The
62 .Bx 4.3
63 option syntax is implemented for backward compatibility but
64 is not documented here.)
65 .Sh DESCRIPTION
66 .Nm Dump
67 examines files
68 on an ext2 filesystem
69 and determines which files
70 need to be backed up. These files
71 are copied to the given disk, tape or other
72 storage medium for safe keeping (see the
73 .Fl f
74 option below for doing remote backups).
75 A dump that is larger than the output medium is broken into
76 multiple volumes.
77 On most media the size is determined by writing until an
78 end-of-media indication is returned.
79 .Pp
80 On media that cannot reliably return an end-of-media indication
81 (such as some cartridge tape drives), each volume is of a fixed size;
82 the actual size is determined by specifying cartridge media, or via the
83 tape size, density and/or block count options below.
84 By default, the same output file name is used for each volume
85 after prompting the operator to change media.
86 .Pp
87 .Ar files-to-dump
88 is either a mountpoint of a filesystem
89 or a list of files and directories to be backed up as a subset of a
90 filesystem.
91 In the former case, either the path to a mounted filesystem
92 or the device of an unmounted filesystem can be used.
93 In the latter case, certain restrictions are placed on the backup:
94 .Fl u
95 is not allowed, the only dump level that is supported is
96 .Fl 0
97 and all the files and directories must reside on the same filesystem.
98 .Pp
99 The following options are supported by
100 .Nm Ns :
101 .Bl -tag -width Ds
102 .It Fl 0\-9
103 Dump levels.
104 A level 0, full backup,
105 guarantees the entire file system is copied
106 (but see also the
107 .Fl h
108 option below).
109 A level number above 0,
110 incremental backup,
111 tells
112 .Nm dump
113 to
114 copy all files new or modified since the
115 last dump of a lower level.
116 The default level is 9.
117 .It Fl B Ar records
118 The number of 1 kB blocks per volume.
119 This option overrides the end-of-media detection, and calculation
120 of tape size based on length and density. If compression is on this
121 limits the size of the compressed output per volume.
122 .It Fl a
123 .Dq auto-size .
124 Bypass all tape length calculations, and write
125 until an end-of-media indication is returned. This works best
126 for most modern tape drives, and is the default.
127 Use of this option is particularly recommended when appending to an
128 existing tape, or using a tape drive with hardware compression
129 (where you can never be sure about the compression ratio).
130 .It Fl b Ar blocksize
131 The number of kilobytes per dump record.
132 Since the IO system slices all requests into chunks of MAXBSIZE
133 (typically 64kB), it is not possible to use a larger blocksize
134 without having problems later with
135 .Xr restore 8 .
136 Therefore
137 .Nm dump
138 will constrain writes to MAXBSIZE.
139 The default blocksize is 10.
140 .It Fl c
141 Change the defaults for use with a cartridge tape drive, with a density
142 of 8000 bpi, and a length of 1700 feet. Specifying a cartridge drive
143 overrides the end-of-media detection.
144 .It Fl e Ar inode
145 Exclude
146 .Ar inode
147 from the dump (you can use
148 .Ar stat
149 to find the inode number for a file or directory).
150 .It Fl h Ar level
151 Honor the user
152 .Dq nodump
153 flag
154 .Dp Dv UF_NODUMP
155 only for dumps at or above the given
156 .Ar level .
157 The default honor level is 1,
158 so that incremental backups omit such files
159 but full backups retain them.
160 .It Fl d Ar density
161 Set tape density to
162 .Ar density .
163 The default is 1600BPI. Specifying a tape density overrides the
164 end-of-media detection.
165 .It Fl f Ar file
166 Write the backup to
167 .Ar file ;
168 .Ar file
169 may be a special device file
170 like
171 .Pa /dev/st0
172 (a tape drive),
173 .Pa /dev/rsd1c
174 (a floppy disk drive),
175 an ordinary file,
176 or
177 .Ql Fl
178 (the standard output).
179 Multiple file names may be given as a single argument separated by commas.
180 Each file will be used for one dump volume in the order listed;
181 if the dump requires more volumes than the number of names given,
182 the last file name will used for all remaining volumes after prompting
183 for media changes.
184 If the name of the file is of the form
185 .Dq host:file
186 or
187 .Dq user@host:file
188 .Nm
189 writes to the named file on the remote host using
190 .Xr rmt 8 .
191 The default path name of the remote
192 .Xr rmt 8
193 program is
194 .\" rmt path, is the path on the remote host
195 .Pa /etc/rmt ;
196 this can be overridden by the environment variable
197 .Ev RMT .
198 .It Fl F Ar script
199 Run script at the end of each tape. The device name and the
200 current volume number are passed on the command line.
201 The script must return 0 if
202 .Nm
203 should continue without asking the user to change the tape, 1 if
204 .Nm
205 should continue but ask the user to change the tape.
206 Any other exit code will cause
207 .Nm
208 to abort.
209 For security reasons,
210 .Nm
211 reverts back to the real user ID and the real group ID before
212 running the script.
213 .It Fl k
214 Use Kerberos authentication to talk to remote tape servers. (Only
215 available if this option was enabled when
216 .Nm
217 was compiled.)
218 .It Fl L Ar label
219 The user-supplied text string
220 .Ar label
221 is placed into the dump header, where tools like
222 .Xr restore 8
223 and
224 .Xr file 1
225 can access it.
226 Note that this label is limited
227 to be at most LBLSIZE (currently 16) characters, which must include
228 the terminating
229 .Ql \e0 .
230 .It Fl M
231 Enable the multi-volume feature. The name specified with
232 .Fl f
233 is treated as a prefix and
234 .Nm
235 writes in sequence to <prefix>001, <prefix>002 etc. This can be
236 useful when dumping to files on an ext2 partition, in order to bypass
237 the 2GB file size limitation.
238 .It Fl n
239 Whenever
240 .Nm
241 requires operator attention,
242 notify all operators in the group
243 .Dq operator
244 by means similar to a
245 .Xr wall 1 .
246 .It Fl Q Ar file
247 Enable the Quick File Access support. Tape positions for each
248 inode are stored into the file
249 .Ar file
250 which is used by restore (if called with parameter Q and the filename)
251 to directly position the tape at the file restore is currently working
252 on. This saves hours when restoring single files from large backups,
253 saves the tapes and the drive's head.
254 .Pp
255 It is recommended to set up the st driver to return logical tape
256 positions rather than physical before calling dump/restore with
257 parameter Q. Since not all tape devices support physical tape
258 positions those tape devices return an error during dump/restore when
259 the st driver is set to the default physical setting.
260 Please see the st man page, option MTSETDRVBUFFER, or the mt man
261 page, on how to set the driver to return logical tape positions.
262 .Pp
263 Before calling restore with parameter Q, always make sure the st
264 driver is set to return the same type of tape position used during the
265 call to dump. Otherwise restore may be confused.
266 .It Fl s Ar feet
267 Attempt to calculate the amount of tape needed at a particular density.
268 If this amount is exceeded,
269 .Nm
270 prompts for a new tape.
271 It is recommended to be a bit conservative on this option.
272 The default tape length is 2300 feet. Specifying the tape size
273 overrides end-of-media detection.
274 .ne 1i
275 .It Fl S
276 Size estimate. Determine the amount of space
277 that is needed to perform the dump without
278 actually doing it, and display the estimated
279 number of bytes it will take. This is useful
280 with incremental dumps to determine how many
281 volumes of media will be needed.
282 .It Fl T Ar date
283 Use the specified date as the starting time for the dump
284 instead of the time determined from looking in
285 .Pa __DUMPDATES__ .
286 The format of
287 .Ar date
288 is the same as that of
289 .Xr ctime 3 .
290 This option is useful for automated dump scripts that wish to
291 dump over a specific period of time.
292 The
293 .Fl T
294 option is mutually exclusive from the
295 .Fl u
296 option.
297 .It Fl u
298 Update the file
299 .Pa __DUMPDATES__
300 after a successful dump.
301 The format of
302 .Pa __DUMPDATES__
303 is readable by people, consisting of one
304 free format record per line:
305 filesystem name,
306 increment level
307 and
308 .Xr ctime 3
309 format dump date.
310 There may be only one entry per filesystem at each level.
311 The file
312 .Pa __DUMPDATES__
313 may be edited to change any of the fields,
314 if necessary.
315 .It Fl W
316 .Nm Dump
317 tells the operator what file systems need to be dumped.
318 This information is gleaned from the files
319 .Pa __DUMPDATES__
320 and
321 .Pa /etc/fstab .
322 The
323 .Fl W
324 option causes
325 .Nm
326 to print out, for all file systems in
327 .Pa __DUMPDATES__ ,
328 and regognized file systems in
329 .Pa /etc/fstab .
330 the most recent dump date and level,
331 and highlights those that should be dumped.
332 If the
333 .Fl W
334 option is set, all other options are ignored, and
335 .Nm
336 exits immediately.
337 .It Fl w
338 Is like
339 .Fl W ,
340 but prints only recognized filesystems in
341 .Pa /etc/fstab
342 which need to be dumped.
343 .It Fl z Ar compression level
344 Compress every block to be written on the tape using zlib library. This
345 option will work only when dumping to a file or pipe or, when dumping
346 to a tape drive, if the tape drive is capable of writing variable
347 length blocks. You will need at least the 0.4b22 version of restore in
348 order to extract compressed tapes. Tapes written using compression will
349 not be compatible with the BSD tape format. The (optional) parameter
350 specifies the compression level zlib will use. The default compression
351 level is 2.
352 .El
353 .Pp
354 .Nm Dump
355 requires operator intervention on these conditions:
356 end of tape,
357 end of dump,
358 tape write error,
359 tape open error or
360 disk read error (if there is more than a threshold of 32).
361 In addition to alerting all operators implied by the
362 .Fl n
363 key,
364 .Nm
365 interacts with the operator on
366 .Em dump's
367 control terminal at times when
368 .Nm
369 can no longer proceed,
370 or if something is grossly wrong.
371 All questions
372 .Nm
373 poses
374 .Em must
375 be answered by typing
376 .Dq yes
377 or
378 .Dq no ,
379 appropriately.
380 .Pp
381 Since making a dump involves a lot of time and effort for full dumps,
382 .Nm
383 checkpoints itself at the start of each tape volume.
384 If writing that volume fails for some reason,
385 .Nm
386 will,
387 with operator permission,
388 restart itself from the checkpoint
389 after the old tape has been rewound and removed,
390 and a new tape has been mounted.
391 .Pp
392 .Nm Dump
393 tells the operator what is going on at periodic intervals,
394 including usually low estimates of the number of blocks to write,
395 the number of tapes it will take, the time to completion, and
396 the time to the tape change.
397 The output is verbose,
398 so that others know that the terminal
399 controlling
400 .Nm
401 is busy,
402 and will be for some time.
403 .Pp
404 In the event of a catastrophic disk event, the time required
405 to restore all the necessary backup tapes or files to disk
406 can be kept to a minimum by staggering the incremental dumps.
407 An efficient method of staggering incremental dumps
408 to minimize the number of tapes follows:
409 .Bl -bullet -offset indent
410 .It
411 Always start with a level 0 backup, for example:
412 .Bd -literal -offset indent
413 /sbin/dump -0u -f /dev/st0 /usr/src
414 .Ed
415 .Pp
416 This should be done at set intervals, say once a month or once every two months,
417 and on a set of fresh tapes that is saved forever.
418 .It
419 After a level 0, dumps of active file
420 systems are taken on a daily basis,
421 using a modified Tower of Hanoi algorithm,
422 with this sequence of dump levels:
423 .Bd -literal -offset indent
424 3 2 5 4 7 6 9 8 9 9 ...
425 .Ed
426 .Pp
427 For the daily dumps, it should be possible to use a fixed number of tapes
428 for each day, used on a weekly basis.
429 Each week, a level 1 dump is taken, and
430 the daily Hanoi sequence repeats beginning with 3.
431 For weekly dumps, another fixed set of tapes per dumped file system is
432 used, also on a cyclical basis.
433 .El
434 .Pp
435 After several months or so, the daily and weekly tapes should get
436 rotated out of the dump cycle and fresh tapes brought in.
437 .Sh ENVIRONMENT
438 .Bl -tag -width Fl
439 .It Ev TAPE
440 If no -f option was specified,
441 .Nm
442 will use the device specified via
443 .Ev TAPE
444 as the dump device.
445 .Ev TAPE
446 may be of the form
447 .Qq tapename ,
448 .Qq host:tapename ,
449 or
450 .Qq user@host:tapename .
451 .It Ev RMT
452 The environment variable
453 .Ev RMT
454 will be used to determine the pathname of the remote
455 .Xr rmt 8
456 program.
457 .It Ev RSH
458 .Nm Dump
459 uses the contents of this variable to determine the name of the
460 remote shell command to use when doing remote backups (rsh, ssh etc.).
461 If this variable is not set,
462 .Xr rcmd 3
463 will be used, but only root will be able to do remote backups.
464 .Sh FILES
465 .Bl -tag -width __DUMPDATES__ -compact
466 .It Pa /dev/st0
467 default tape unit to dump to
468 .It Pa __DUMPDATES__
469 dump date records
470 .It Pa /etc/fstab
471 dump table: file systems and frequency
472 .It Pa /etc/group
473 to find group
474 .Em operator
475 .El
476 .Sh SEE ALSO
477 .Xr fstab 5 ,
478 .Xr restore 8 ,
479 .Xr rmt 8
480 .Sh DIAGNOSTICS
481 Many, and verbose.
482 .Pp
483 .Nm Dump
484 exits with zero status on success.
485 Startup errors are indicated with an exit code of 1;
486 abnormal termination is indicated with an exit code of 3.
487 .Sh BUGS
488 It might be considered a bug that this version of dump can only handle ext2
489 filesystems. Specifically, it does not work with FAT filesystems.
490 .Pp
491 Fewer than 32 read errors on the filesystem are ignored. If noticing
492 read errors is important, the output from dump can be parsed to look for lines
493 that contain the text 'read error'.
494 .Pp
495 Each reel requires a new process, so parent processes for
496 reels already written just hang around until the entire tape
497 is written.
498 .Pp
499 The estimated number of tapes is not correct if compression is on.
500 .Pp
501 It would be nice if
502 .Nm
503 knew about the dump sequence,
504 kept track of the tapes scribbled on,
505 told the operator which tape to mount when,
506 and provided more assistance
507 for the operator running
508 .Xr restore .
509 .Pp
510 .Nm Dump
511 cannot do remote backups without being run as root, due to its
512 security history.
513 Presently, it works if you set it setuid (like it used to be), but this
514 might constitute a security risk. Note that you can set RSH to use
515 a remote shell program instead.
516 .Sh AUTHOR
517 The
518 .Nm dump/restore
519 backup suite was ported to Linux's Second Extended File System
520 by Remy Card <card@Linux.EU.Org>. He maintained the initial versions
521 of dump (up and including 0.4b4, released in january 1997).
522 .Pp
523 Starting with 0.4b5, the new maintainer is Stelian Pop
524 .br
525 <pop@noos.fr>.
526 .Sh AVAILABILITY
527 The
528 .Nm dump/restore
529 backup suite is available from
530 .br
531 http://dump.sourceforge.net
532 .Sh HISTORY
533 A
534 .Nm
535 command appeared in
536 .At v6 .