UNIX for Dummies Questions & Answers

I was at a vendor demonstration last week, and the presenter had his laptop on screen where we could see the product demo. He had a Unix terminal screen up and created an empty file, but he used some command (I wish I had caught how he did this) to actually make the empty file a certain size. Does anyone know how this is accomplished? In this instance, he used it to show how a 20 Mb file would be transferred through a router.


Anyone know how this was possibly done?

i think the 'dd' command can do this with the bs=BYTES, obs=BYTES, and of=FILE arguments.

Some systems have a prealloc command. HP-UX does, SunOS does not. So it's not universal.

I think the trick is to produce the file, say 20MB, without a 20MB input file. Off hand, I can't do this with dd. I tried to cat /dev/null into dd, i.e.

Code:

cat /dev/null | dd of=test.file bs=20480000 (well almost!)

but this will not work because nothing comes in dd to count

If I had a 20MB file, I could pipe that into dd, but then if I had a 20MB file, what's the point?

Some systems have a /dev/zero that produces a infinite number of zeros. Or a /dev/full that produces a infinite amount of garbage. You need to use something like that.

Perfect, something like:

Code:

cat /dev/zero | dd of=test.file bs=20000

Thanks!

Not sure how to make 20MBs from this Perhaps someone can assist? I tried bs=20000 but it was HUGH!!! Much bigger than 20MB!!!!

Neo

what about a while() loop?
something like:

Code:

while ($filesize != 2000000)
{cat somefile of a managable size  > newfile}

now this dosnt create a file out of nothing, but put inside of a small script, well theres your arbitrary file size creation utility.


EDIT: didnt see perderabos and neos follow up posts, so maybe you could 'cat /dev/zero' in a while loop...