1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
|
dds2tar Digital-Data-Storage Extract Tool for LINUX and HP-UX
This tool makes use of the fast seek command of DAT devices.
Files from a selected file archive can be extracted within
one minute.
This tool was originally written for Linux SCSI tape archives.
All device dependent code is separated.
It should be easy, to change this for your machine type.
The only problem should be the ioctls for
MTIOCTOP,(MTSEEK,arg) and MTIOCPOS.
Les.Grant@mitsuibabcock.co.uk send me a patch to make it
working under HP-UX, thanks.
Other patches epsecially for Solaris are welcome.
J"org Weule weule@cs.uni-duesseldorf.de
Fon: +49 211 751409
This software is available at
ftp.uni-duesseldorf.de:/pub/unix/apollo
( This is my personal directory, I started writing PD-software
on Apollo-Computers, but they are dying now :-( . Now you
will find some other stuff there.)
----------------------------------------------------------------------------
LICENSE
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published
by the Free Software Foundation; either version 1, or (at your option)
any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
----------------------------------------------------------------------------
DESCRIPTION
To extract a file from a tape, you have to know the location of the file.
A tar archive is a sequence of tar records, 512 byte each. The blocksize
(tar -b #) selects the number of tar records inside each block of the tape.
(HP is calling the blocks of the tape records also.)
The location of a file is a tuple of the block number and the number of the
record where the header of the file can be found inside the block.
The first step will be to extract a table with block and record information
of all files in one archive. If the archive is the 10'th file of the tape,
you will get the table by:
mt asf 10
dds2index -t my-archive-table -f /dev/nst0
Tar is writing some record information to stdout, when you select the options
-R and -v. To use this information as a location table, you have to know the
block number of the first block and the blocksize of the archive. If it's the
first archive of the tape and the blocksize is the default of 20, you can
use the verbose lines of tar as a location table. (Additional lines with
warnings can be included). Otherwise you have to patch tar to retrieve the
location information. If you don't want to patch tar, feel free to insert
the output of 'mt-dds tell' at the top of the verbose output of tar.
If you have applied the patch file 'tar-1.12.patch' to GNU tar, you can use
the verbose output of the tar command (tar -c -v --record-file F) instead of the
index file generated with the dds2index command above.
(If you patch tar-1.12, you are able to write the record-information
to a file while getting the errors on screen. The patch 'tar-1.12.patch'
adds the option '--record-file F'. Additionally a small bug in the source file
'create.c' of tar is fixed. I have moved one line to correct the verbose output
of the record numbers. While creating the archive, you can store the current
time at the label record. A bug fix to write this date to the verbose output
is also done in the source file 'list.c'.)
! The patch tar-1.12.patch moves one line inside the source file 'create.c'
! and adds some lines for the ability to direct the verbose output to a file.
! Two extra lines are printed to the terminal in some cases to get to current
! tape position at the start and the blocksize.
! Nothing else is done. There should be no extra risk for your data.
! The tar will stay fully compatible to the original version.
! The patch is also convenient for other operating systems.
Lines of the index file stating with '#' are ignored. So you can insert
some comments. This distribution includes some examples of index files.
Look for files 'index-of-*'
Once you have the table, you can get the files very quickly. Lets suppose,
you want to extract everything of the glibc, you have to invoke
the following command:
dds2tar -t my-archive-table 'glibc*' | tar -f - -xvv
dds2tar will search in the archive table for the blocks to read, and
extracts the records with the selected files. These records together
are giving a valid tar archive, to use tar for restoring to your disk.
This will also work, if the files are not a continuous area of the tape.
dds2tar is skipping over the gaps very quickly with a seek command.
If you want to extract exactly one file, use -x instead of -p.
If you want to extract all files, witch named include a pattern,
use -g instead of -p or -x.
A list of files is valid to search for.
With the option -z, you can force the index file to be in compressed mode.
Compressed archive files are not supported.
If you don't use the option -t, stdin or stdout are used in this case.
The default device is /dev/nst0, witch can be overwritten by the TAPE variable
and the option -f.
If you want to get familiar with the command, use
# skipping to some file
mt rewind
mt fsf 20
# creating the table of the archive
dds2index -t my-archive-table
# The following line is an alternative way to create the index.
# You have to patch tar to make the option --record-file
# available.
# tar -t --record-file my-archive-table
# extract all files with the substring glibc in there names
# and use tar to check the output of dds2tar.
dds2tar -t my-archive-table 'glibc*' | tar -f - -tvv
# See what's going on ... ^^^^ ^^
to see what's going on. In any case, dds2tar is reading your tape,
so don't worry. Only the file specified by -t is written by this example.
NOTE:
To make extraction easier, dds2tar will also extract the directories
for the selected files. So you will also get the full directory hierarchy.
The owners of HP-DAT devices will find the actions comp-on, comp-off and
comp-query (dds-compression-mode) of mt-dds convenient to change the compression
mode of the device. Turning off the compression will speed up storing
'gzip'ed files.
WARNING:
If we seek to the first block of a file, it can happen, that we are
going to the position before the filemark. The program is closing and opening
the file in that case to step over the filemark. Don't use /dev/st0 in that
case, use /dev/nst0.
THANKS:
Thanks to Andreas (Andreas_Bagge@h2.maus.de), who has send me a new manual
page, witch gave me the ideas how to split dds2tar-1.1.3 into peaces.
His page was also the starting point for the approved man-page added to this
version.
Thanks also to Chris Hanson (cph@martigny.ai.mit.edu) for his bug bix.
Last change 2.2.95 by J"org Weule
----------------------------------------------------------------------------
INSTALLATION
1. Change the "Makefile" to reflect the buffersize of the kernel.
Otherwise uncomment the definition of BUFFER.
2. type "make install".
3. patch tar-1.12 if you like.
|
