dd

Convert and copy a file

Supplemental Information

The dd command is used to copy a file while performing conversions and formatting on the original content. It is a powerful tool often used for low-level operations where other commands might not suffice. A common use case is backing up raw devices, though for specific applications like databases, specialized tools are generally preferred for better manageability.

It is recommended to use dd for physical disk operations. For file system tasks, commands like tar or cpio are often more convenient. When using dd for disk operations, it is best to use block device files.

Syntax

dd (options)

Options

bs=<bytes>: Set both input (ibs) and output (obs) block size to the specified number of bytes;
cbs=<bytes>: Convert the specified number of bytes at a time;
conv=<keywords>: Specify how the file should be converted (comma-separated list);
count=<blocks>: Copy only the specified number of input blocks;
ibs=<bytes>: Read the specified number of bytes at a time;
obs=<bytes>: Write the specified number of bytes at a time;
if=<file>: Read from the specified file instead of stdin;
of=<file>: Write to the specified file instead of stdout;
seek=<blocks>: Skip the specified number of blocks at the start of the output;
skip=<blocks>: Skip the specified number of blocks at the start of the input;
--help: Display help;
--version: Display version information.

Examples

[root@localhost text]# dd if=/dev/zero of=sun.txt bs=1M count=1
1+0 records in
1+0 records out
1048576 bytes (1.0 MB) copied, 0.006107 seconds, 172 MB/s

[root@localhost text]# du -sh sun.txt 
1.1M    sun.txt

This command creates a 1MB file named sun.txt. Parameter explanation:

if: Input file. Defaults to stdin if not specified.
of: Output file. Defaults to stdout if not specified.
bs: Block size in bytes.
count: Number of blocks to copy.
/dev/zero: A character device that provides as many null bytes (\0) as requested.

Block size units:

Unit	Code
Byte (1B)	c
Word (2B)	w
Block (512B)	b
Kilobyte (1024B)	k
Megabyte (1024KB)	M
Gigabyte (1024MB)	G

The output also shows the performance of the operation:

1048576 bytes (1.0 MB) copied, 0.006107 seconds, 172 MB/s

Generating a Random String

You can use /dev/urandom with dd to generate random strings:

[root@localhost ~]# dd if=/dev/urandom bs=1 count=15 | base64 -w 0
15+0 records in
15+0 records out
15 bytes (15 B) copied, 0.000111993 s, 134 kB/s
wFRAnlkXeBXmWs1MyGEs

Testing Disk Write Speed

[root@localhost ~]# dd if=/dev/zero of=/tmp/testfile bs=1G count=1 oflag=direct
1+0 records in
1+0 records out
1073741824 bytes (1.1 GB) copied, 7.10845 s, 151 MB/s

Testing Disk Read Speed

[root@localhost ~]# dd if=/tmp/testfile of=/dev/null bs=1G count=1 iflag=direct
1+0 records in
1+0 records out
1073741824 bytes (1.1 GB) copied, 6.53009 s, 164 MB/s