28

What is a quick and easy way to make a file that is, say, 2 GB in size?

ændrük
  • 76,794

3 Answers3

40

The zero-fill method (here modified to avoid potential memory bottlenecks) took 17 seconds to create a 10 GB file on an SSD and caused Ubuntu's graphical interface to become unresponsive.

$ time sh -c 'dd if=/dev/zero iflag=count_bytes count=10G bs=1M of=large; sync'
10240+0 records in
10240+0 records out
10737418240 bytes (11 GB, 10 GiB) copied, 17.2003 s, 624 MB/s

real    0m17.642s
user    0m0.008s
sys     0m9.404s
$ du -B 1 --apparent-size large
10737418240     large
$ du -B 1 large
10737422336     large

fallocate creates large files instantly by directly manipulating the file's allocated disk space:

$ time sh -c 'fallocate -l 10G large; sync'

real    0m0.038s
user    0m0.000s
sys     0m0.016s
$ du -B 1 --apparent-size large
10737418240     large
$ du -B 1 large
10737422336     large

truncate also works instantly, and creates sparse files which don't use up actual disk space until data is written later on:

$ time sh -c 'truncate -s 10G large; sync'

real    0m0.014s
user    0m0.000s
sys     0m0.004s
$ du -B 1 --apparent-size large
10737418240     large
$ du -B 1 large
0       large
ændrük
  • 76,794
  • dd ... bs=2G count=1 reads 2 GB into memory (in one read(2) call). If you've got memory pressure that's probably not the way to go. More, smaller blocks may be faster if it means less paging. – claymation Apr 04 '17 at 05:34
22

An easy way would be to use the dd command to write a file full of zeros. 

dd if=/dev/zero of=outputFile bs=2G count=1
  • if = input file
  • of = output file
  • bs = bytes

Use G in the size argument if you want computer (1024*1024*1024) gigabytes, or GB if you want human (1000*1000*1000) gigabytes.

Ward Muylaert
  • 3,882
MikeVB
  • 393
  • 1
    i will just add that if you dont want all zeros, you can choose if=/dev/random – Denwerko Jun 10 '11 at 06:28
  • Though /dev/random will take more time. – con-f-use Jun 10 '11 at 08:11
  • 10
    Using /dev/random will take an awful lot of time. Use /dev/urandom in that case (it's non-blocking, but not guaranteed to have the same level of randomness). Drawing 2 GB from either one will almost certainly completely exhaust your system's entropy, so don't do anything cryptographic for a while afterwards. – user Jun 10 '11 at 08:39
1

ftp://ftp.fsf.hu/testfiles/maketestfiles.sh

or Seek is the size of the file you want in bytes - 1.

dd if=/dev/zero of=filename.big bs=1 count=1 seek=1048575 # 1 MByte

dd if=/dev/zero of=filename.big bs=1 count=1 seek=10485759 # 10 MByte

dd if=/dev/zero of=filename.big bs=1 count=1 seek=104857599 # 100 MByte

dd if=/dev/zero of=filename.big bs=1 count=1 seek=1073741823 # 1024 MByte

dd if=/dev/zero of=filename.big bs=1 count=1 seek=42949672959 # 40960 MByte
LanceBaynes
  • 1,055