ddd - double-speed data dumper
ddd [option=value] ...
Ddd works almost the same way as dd(1), but it has a much better throughput, especially when used with slow i/o-devices, such as tape
drives. The improvement is achieved mainly by dividing the copying process into two processes, one of which reads while the other one
writes and vice versa. Also all code conversion capabilities are omitted. There is no additional overhead copying data between various
Ddd was inspired by the vast difference in speed between BSD4.2 and BSD4.3 dumps - in BSD4.3 dump(8) uses alternating processes to write to
raw magnetic tape, thus keeping the tape continuously in motion. I wanted to get the same improvement to remote dumps, so this filter was
needed. Directing all physical I/O through ddd usually increases the throughput of any pipeline of unix commands (if you have enough MIPS
and RAM to handle two extra processes).
Ddd uses options if, of, ibs and obs exactly as dd(1). Option bs can also be used to specify ibs and obs at once. One option differs
slightly in meaning: cbs can be used to specify the size of the internal buffer. Input and output processes will swap duties when cbs
bytes have been transferred. Default values for all sizes are 512 bytes. As with dd(1), letters k (kilobyte), b (block) or w (word) can
be appended to size values. Other options are not provided.
For best performance, block sizes should be rather large. For magnetic tape, I use obs=100b and cbs=500b or so. Large block sizes (~100b)
are also effective for network connections. However, cbs should be small enough for all the data to fit in core, since page faults add
Tapani Lindgren <email@example.com>
Laboratory of Information Processing Science
Helsinki University of Technology
dd(1), tar(1), dump(8)
Should you find one, let me know!
(Applies to U.S. residents & citizens only)
Do not use this program! Get rid of it as soon as you can! It will probably corrupt all your data, break down your computer and cause
severe injury to the operators. Even reading the source code may give you a headache. I warned you! I will take no responsibility what-
Check Out this Related Man Page
dd(1) General Commands Manual dd(1)
dd - Converts and copies a file
Interfaces documented on this reference page conform to industry standards as follows:
dd: XCU5.0, SVID 4
Refer to the standards(5) reference page for more information about industry standards and associated tags.
The option=value operand set may take any of the following forms: Specifies the input file name; standard input is the default. Specifies
the output file name; standard output is the default. For the above operands, the application must supply an expression specifying a size
in bytes. The expression, expr can be: a positive decimal number a positive decimal number followed by k specifying mutiplication by 1024 a
positive decimal number followed by b specifying multiplication by 512 two or more positive decimal numbers with or without k or b seper-
ated by x, specifying the product of the indicated values.
All the operands will be processed before any input is read. Skips number input records before starting copy. [Tru64 UNIX] Copies
number input files before terminating (makes sense only where input is a magnetic tape or similar device). [Tru64 UNIX] Seeks to
the numberth record from the beginning of input file before copying. [Tru64 UNIX] Seeks to the numberth record from the beginning
of output file before copying. Same as seek=number. Seeks to the numberth record from the beginning of output file before copying.
Same as oseek=number. Copies only number input records. Specifies one or more of the following conversions: Converts EBCDIC to
ASCII. Converts variable-length records to fixed-length. Converts ASCII to EBCDIC. Converts IBM-EBCDIC to ASCII. Performs a
slightly different map of ASCII to EBCDIC. Converts fixed-length records to variable-length. Makes all alphabetic characters lower
case. Makes all alphabetic characters upper case. Swaps every pair of bytes. Does not stop processing on an error. Pads every
input record to ibs. [Tru64 UNIX] Creates a sparse output file as described in AdvFS Administration. Do not truncate the output
file. Preserve blocks in the output file not explicitly written by this invocation of the dd utility. (See the of=output_file oper-
and.) Allows several comma-separated conversions.
The dd command reads the specified input file or standard input, does the specified conversions, and copies it to the specified output file
or standard output. The input and output block size may be specified to take advantage of raw physical I/O. The terms block and record
refer to the quantity of data read or written by dd in one operation and are not necessarily the same size as a disk block.
Where sizes are specified, a number of bytes is expected. A number may end with w, b, or k to specify multiplication by 2, 512, or 1024,
respectively; a pair of numbers can be separated by an x to indicate a product.
The cbs specification is used if one of the following conversions is specified: ascii, unblock, ebcdic, ibm, or block. For the first two
conversions, dd places characters in a conversion buffer of size cbs, converts these characters to ASCII, trims trailing spaces, and adds
newline characters before sending data to the specified output. For the latter three cases, dd places ASCII characters in the conversion
buffer, converts these characters to EBCDIC, and adds trailing spaces to create an output record of size cbs.
After it finishes, dd reports the number of whole and partial input and output blocks.
[Tru64 UNIX] To copy to a raw disk, the disk label must first be zeroed using the disklabel -z command. For example: disklabel -z
[Tru64 UNIX] If you do not zero out the disk label, dd fails with the following error message: dd write error: Read-only file sys-
tem The ASCII/EBCDIC conversion tables are taken from the 256 character standard in the CACM November, 1968. There is no universal
standard for EBCDIC/ASCII translation. [Tru64 UNIX] One must specify conv=noerror,sync when copying raw disks with bad sectors to
ensure dd stays synchronized. [Tru64 UNIX] Certain combinations of arguments to conv= are permitted. However, the block or
unblock option cannot be combined with ascii, ebcdic, or ibm. Invalid combinations silently ignore all but the last mutually exclu-
sive keyword. [Tru64 UNIX] If you need to use dd to copy to a streaming tape and the data is an odd length (not a multiple of 512
bytes), you must use the conv=sync flag to fill the last record. Streaming tape devices permit only multiples of 512 bytes. [Tru64
UNIX] If option bs is used (or bs is equal to obs) and no conversion is specified, then dd is particularly efficient since less
memory copies are done. [Tru64 UNIX] The dd command does not support floppy disk multivolumes, but it does support tape multivol-
umes. This means that when ENOSPC is returned while reading or writing a tape, dd will prompt the user for a new tape.
[Tru64 UNIX] In order to make use of tape multivolumes, the files option must be used.
[Tru64 UNIX] Any file system archive that contains ACLs (access control lists) that was created using dd is not exportable unless the tar-
get system has the exact same password and group files. If there is a mismatch, incorrect access may be granted to a file or directory.
The following exit values are returned: The input file was successfully copied. An error occurred
This message specifies the number of full and partial records both read and written:
f+p records in f+p records out
The number of full records read or written (f) refers to the blocks of data of size ibs or obs. The number of partial records read or
written (p) refers to the blocks of data smaller than ibs or obs.
To read an EBCDIC tape blocked ten 80-byte EBCDIC card images per record into the ASCII file x, enter: dd if=/dev/tape/tape0_d0 of=x
ibs=800 cbs=80 conv=ascii,lcase
Note the use of raw magnetic tape. The dd command is especially suited to I/O on the raw physical devices because it allows reading
and writing in arbitrary record sizes. To convert an ASCII text file to EBCDIC, enter: dd if=text.ascii of=text.ebcdic conv=ebcdic
This converts text.ascii to EBCDIC representation, storing this in text.ebcdic.
The following environment variables affect the execution of dd: Provides a default value for the internationalization variables that are
unset or null. If LANG is unset or null, the corresponding value from the default locale is used. If any of the internationalization vari-
ables contain an invalid setting, the utility behaves as if none of the variables had been defined. If set to a non-empty string value,
overrides the values of all the other internationalization variables. Determines the locale for the interpretation of sequences of bytes
of text data as characters (for example, single-byte as opposed to multibyte characters in arguments). Determines the locale for the for-
mat and contents of diagnostic messages written to standard error. Determines the location of message catalogues for the processing of
Commands: cp(1), cpio(1), sed(1), tar(1), tr(1), trbsd(1)
Command and Shell User's Guide