xref: /netbsd/sbin/dump_lfs/dump_lfs.8 (revision bf9ec67e)
1.\"	$NetBSD: dump_lfs.8,v 1.6 2002/01/21 18:15:08 wiz Exp $
2.\"
3.\" Copyright (c) 1980, 1991, 1993
4.\"	 Regents of the University of California.
5.\" All rights reserved.
6.\"
7.\" Redistribution and use in source and binary forms, with or without
8.\" modification, are permitted provided that the following conditions
9.\" are met:
10.\" 1. Redistributions of source code must retain the above copyright
11.\"    notice, this list of conditions and the following disclaimer.
12.\" 2. Redistributions in binary form must reproduce the above copyright
13.\"    notice, this list of conditions and the following disclaimer in the
14.\"    documentation and/or other materials provided with the distribution.
15.\" 3. All advertising materials mentioning features or use of this software
16.\"    must display the following acknowledgement:
17.\"	This product includes software developed by the University of
18.\"	California, Berkeley and its contributors.
19.\" 4. Neither the name of the University nor the names of its contributors
20.\"    may be used to endorse or promote products derived from this software
21.\"    without specific prior written permission.
22.\"
23.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33.\" SUCH DAMAGE.
34.\"
35.\"     @(#)dump.8	8.3 (Berkeley) 5/1/95
36.\"
37.Dd September 3, 1999
38.Dt DUMP_LFS 8
39.Os
40.Sh NAME
41.Nm dump_lfs ,
42.Nm rdump_lfs
43.Nd filesystem backup
44.Sh SYNOPSIS
45.Nm ""
46.Op Fl 0123456789cnSu
47.Op Fl B Ar records
48.Op Fl b Ar blocksize
49.Op Fl d Ar density
50.Op Fl f Ar file
51.Op Fl h Ar level
52.Op Fl k Ar read blocksize
53.Op Fl L Ar label
54.Op Fl r Ar cachesize
55.Op Fl s Ar feet
56.Op Fl T Ar date
57.Ar files-to-dump
58.Nm ""
59.Op Fl W Li \&| Fl w
60.Pp
61.in -\n(iSu
62(The
63.Bx 4.3
64option syntax is implemented for backward compatibility, but
65is not documented here).
66.Sh DESCRIPTION
67.Nm
68examines files on a filesystem and determines which files need to
69be backed up.
70These files are copied to the given disk, tape or other storage
71medium for safe keeping (see the
72.Fl f
73option below for doing remote backups).
74A dump that is larger than the output medium is broken into
75multiple volumes.
76On most media the size is determined by writing until an
77end-of-media indication is returned.
78On media that cannot reliably return an end-of-media indication
79(such as some cartridge tape drives)
80each volume is of a fixed size;
81the actual size is determined by the tape size and density and/or
82block count options below.
83By default, the same output file name is used for each volume
84after prompting the operator to change media.
85.Pp
86.Ar files-to-dump
87is either a mountpoint of a filesystem,
88or a list of files and directories on a single filesystem to be backed
89up as a subset of the filesystem.
90In the former case, either the path to a mounted filesystem,
91or the device of an unmounted filesystem can be used.
92In the latter case, certain restrictions are placed on the backup:
93.Fl u
94is ignored, the only dump level that is supported is
95.Fl 0 ,
96and all of the files must reside on the same filesystem.
97.Pp
98The following options are supported by
99.Nm "" :
100.Bl -tag -width Ds
101.It Fl 0\-9
102Dump levels.
103A level 0, full backup,
104guarantees the entire file system is copied
105(but see also the
106.Fl h
107option below).
108A level number above 0,
109incremental backup,
110tells dump to
111copy all files new or modified since the
112last dump of a lower level.
113The default level is 9.
114.It Fl B Ar records
115The number of kilobytes per volume, rounded
116down to a multiple of the blocksize.
117This option overrides the calculation of tape size
118based on length and density.
119.It Fl b Ar blocksize
120The number of kilobytes per dump record.
121.It Fl c
122Modify the calculation of the default density and tape size to be more
123appropriate for cartridge tapes.
124.It Fl d Ar density
125Set tape density to
126.Ar density .
127The default is 1600 Bits Per Inch (BPI).
128.It Fl f Ar file
129Write the backup to
130.Ar file ;
131.Ar file
132may be a special device file
133like
134.Pa /dev/rst0
135(a tape drive),
136.Pa /dev/rsd1c
137(a disk drive),
138an ordinary file,
139or
140.Ql Fl
141(the standard output).
142Multiple file names may be given as a single argument separated by commas.
143Each file will be used for one dump volume in the order listed;
144if the dump requires more volumes than the number of names given,
145the last file name will used for all remaining volumes after prompting
146for media changes.
147If the name of the file is of the form
148.Qq host:file ,
149or
150.Qq user@host:file ,
151.Nm
152writes to the named file on the remote host using
153.Xr rmt 8 .
154Note that methods more secure than
155.Xr rsh 1
156.Pq such as Xr ssh 1
157can be used to invoke
158.Xr rmt 8
159on the remote host, via the environment variable
160.Ev RCMD_CMD .
161See
162.Xr rcmd 3
163for more details.
164.It Fl h Ar level
165Honor the user
166.Qq nodump
167flag
168.Pq Dv UF_NODUMP
169only for dumps at or above the given
170.Ar level .
171The default honor level is 1,
172so that incremental backups omit such files
173but full backups retain them.
174.It Fl k Ar read blocksize
175The size in kilobyte of the read buffers, rounded up to a multiple of the
176filesystem block size. Default is 32k.
177.It Fl L Ar label
178The user-supplied text string
179.Ar label
180is placed into the dump header, where tools like
181.Xr restore 8
182and
183.Xr file 1
184can access it.
185Note that this label is limited
186to be at most LBLSIZE (currently 16) characters, which must include
187the terminating
188.Ql \e0 .
189.It Fl n
190Whenever
191.Nm
192requires operator attention,
193notify all operators in the group
194.Qq operator
195by means similar to a
196.Xr wall 1 .
197.It Fl r Ar cachesize
198Use that many buffers for read cache operations.
199A value of zero disables the read cache altogether, higher values
200improve read performance by reading larger data blocks from the
201disk and maintaining them in an LRU cache. See the
202.Fl k
203option for the size of the buffers. Maximum is 512, the size of the cache is
204limited to 15% of the avail RAM by default.
205.It Fl s Ar feet
206Attempt to calculate the amount of tape needed
207at a particular density.
208If this amount is exceeded,
209.Nm
210prompts for a new tape.
211It is recommended to be a bit conservative on this option.
212The default tape length is 2300 feet.
213.It Fl S
214Display an estimate of the backup size and the number of tapes
215required, and exit without actually performing the dump.
216.It Fl T Ar date
217Use the specified date as the starting time for the dump
218instead of the time determined from looking in
219.Pa /etc/dumpdates .
220The format of date is the same as that of
221.Xr ctime 3 .
222This option is useful for automated dump scripts that wish to
223dump over a specific period of time.
224The
225.Fl T
226option is mutually exclusive from the
227.Fl u
228option.
229.It Fl u
230Update the file
231.Pa /etc/dumpdates
232after a successful dump.
233The format of
234.Pa /etc/dumpdates
235is readable by people, consisting of one
236free format record per line:
237filesystem name,
238increment level
239and
240.Xr ctime 3
241format dump date.
242There may be only one entry per filesystem at each level.
243The file
244.Pa /etc/dumpdates
245may be edited to change any of the fields,
246if necessary.
247If a list of files or subdirectories is being dumped
248(as opposed to and entire filesystem), then
249.Fl u
250is ignored.
251.It Fl W
252.Nm
253tells the operator what file systems need to be dumped.
254This information is gleaned from the files
255.Pa /etc/dumpdates
256and
257.Pa /etc/fstab .
258The
259.Fl W
260option causes
261.Nm
262to print out, for each file system in
263.Pa /etc/dumpdates
264the most recent dump date and level,
265and highlights those file systems that should be dumped.
266If the
267.Fl W
268option is set, all other options are ignored, and
269.Nm
270exits immediately.
271.It Fl w
272Is like W, but prints only those filesystems which need to be dumped.
273.El
274.Pp
275If
276.Nm
277honors the
278.Qq nodump
279flag
280.Pq Dv UF_NODUMP ,
281files with the
282.Qq nodump
283flag will not be backed up. If a directory has the
284.Qq nodump
285flag, this directory and any file or directory under it will not be backed up.
286.Pp
287.Nm
288requires operator intervention on these conditions:
289end of tape,
290end of dump,
291tape write error,
292tape open error or
293disk read error (if there are more than a threshold of 32).
294In addition to alerting all operators implied by the
295.Fl n
296option,
297.Nm
298interacts with the operator on
299.Nm "" Ns 's
300control terminal at times when
301.Nm
302can no longer proceed,
303or if something is grossly wrong.
304All questions
305.Nm
306poses
307.Em must
308be answered by typing
309.Qq yes
310or
311.Qq no ,
312appropriately.
313.Pp
314Since making a dump involves a lot of time and effort for full dumps,
315.Nm
316checkpoints itself at the start of each tape volume.
317If writing that volume fails for some reason,
318.Nm
319will,
320with operator permission,
321restart itself from the checkpoint
322after the old tape has been rewound and removed,
323and a new tape has been mounted.
324.Pp
325.Nm
326tells the operator what is going on at periodic intervals,
327including usually low estimates of the number of blocks to write,
328the number of tapes it will take, the time to completion, and
329the time to the tape change.
330The output is verbose,
331so that others know that the terminal
332controlling
333.Nm
334is busy,
335and will be for some time.
336.Pp
337In the event of a catastrophic disk event, the time required
338to restore all the necessary backup tapes or files to disk
339can be kept to a minimum by staggering the incremental dumps.
340An efficient method of staggering incremental dumps
341to minimize the number of tapes follows:
342.Bl -bullet -offset indent
343.It
344Always start with a level 0 backup, for example:
345.Bd -literal -offset indent
346/sbin/dump -0u -f /dev/nrst1 /usr/src
347.Ed
348.Pp
349This should be done at set intervals, say once a month or once every two months,
350and on a set of fresh tapes that is saved forever.
351.It
352After a level 0, dumps of active file
353systems are taken on a daily basis,
354using a modified Tower of Hanoi algorithm,
355with this sequence of dump levels:
356.Bd -literal -offset indent
3573 2 5 4 7 6 9 8 9 9 ...
358.Ed
359.Pp
360For the daily dumps, it should be possible to use a fixed number of tapes
361for each day, used on a weekly basis.
362Each week, a level 1 dump is taken, and
363the daily Hanoi sequence repeats beginning with 3.
364For weekly dumps, another fixed set of tapes per dumped file system is
365used, also on a cyclical basis.
366.El
367.Pp
368After several months or so, the daily and weekly tapes should get
369rotated out of the dump cycle and fresh tapes brought in.
370.Pp
371If
372.Nm
373receives a
374.Dv SIGINFO
375signal
376(see the
377.Qq status
378argument of
379.Xr stty 1 )
380whilst a backup is in progress, statistics on the amount completed,
381current transfer rate, and estimated finished time, will be written
382to the standard error output.
383.Sh ENVIRONMENT
384If the following environment variables exist, they are utilized by
385.Nm "" .
386.Bl -tag -width Fl
387.It Ev TAPE
388If no -f option was specified,
389.Nm
390will use the device specified via
391.Ev TAPE
392as the dump device.
393.Ev TAPE
394may be of the form
395.Qq tapename ,
396.Qq host:tapename ,
397or
398.Qq user@host:tapename .
399.It Ev RCMD_CMD
400.Nm
401will use
402.Ev RCMD_CMD
403rather than
404.Xr rsh 1
405to invoke
406.Xr rmt 8
407on the remote machine.
408.El
409.Sh FILES
410.Bl -tag -width /etc/dumpdates -compact
411.It Pa /dev/nrst0
412default tape unit to use. Taken from
413.Dv _PATH_DEFTAPE
414in
415.Pa /usr/include/paths.h .
416.It Pa /dev/rst*
417raw SCSI tape interface
418.It Pa /etc/dumpdates
419dump date records
420.It Pa /etc/fstab
421dump table: file systems and frequency
422.It Pa /etc/group
423to find group
424.Em operator
425.El
426.Sh DIAGNOSTICS
427Many, and verbose.
428.Pp
429.Nm
430exits with zero status on success.
431Startup errors are indicated with an exit code of 1;
432abnormal termination is indicated with an exit code of 3.
433.Sh SEE ALSO
434.Xr chflags 1 ,
435.Xr stty 1 ,
436.Xr fts 3 ,
437.Xr st 4 ,
438.Xr fstab 5 ,
439.Xr restore 8 ,
440.Xr rmt 8
441.Sh HISTORY
442A
443.Nm
444command appeared in
445.Nx 1.5 .
446.Sh BUGS
447Fewer than 32 read errors on the filesystem are ignored.
448.Pp
449Each reel requires a new process, so parent processes for
450reels already written just hang around until the entire tape
451is written.
452.Pp
453.Nm
454with the
455.Fl W
456or
457.Fl w
458options does not report filesystems that have never been recorded
459in
460.Pa /etc/dumpdates ,
461even if listed in
462.Pa /etc/fstab .
463.Pp
464When dumping a list of files or subdirectories, access privileges are
465required to scan the directory (as this is done via the
466.Xr fts 3
467routines rather than directly accessing the filesystem).
468.Pp
469It would be nice if
470.Nm
471knew about the dump sequence,
472kept track of the tapes scribbled on,
473told the operator which tape to mount when,
474and provided more assistance
475for the operator running
476.Xr restore 8 .
477