Need to delete filename from all subdirectories in Linux

I have a single filename that exists thousands of times in the subdirectory tree, and I need to delete it everywhere it exists.

Easy enough in DOS:

delete /s myfile.txt

A similar approach in Linux doesn't work:

rm -r myfile.txt

because it will only remove from the current directory. What is the best approach?
Who is Participating?
woolmilkporcConnect With a Mentor Commented:
That's a task for "find":

find /start/dir -type f -name myfile.txt | xargs rm

find searches all subdirectories according to the given criteria. Results are passed to xargs which in turn calls rm with a number of parameters (= files to delete) suitable for this command.


jbbarnesAuthor Commented:
Hi, I can see how that would work. Thanks.

I do get an error, though, when I try it:

rm: missing operand

The xargs man page gives an example along these lines:

find /start/dir -type f -name myfile.txt -print | xargs rm -f
koffuConnect With a Mentor Commented:
in previous post command is valid, but with space after "xargs rm ". Also, if your filename will have spaces or another special symbols - it will not work.
find /start/dir -type f -name myfile.txt -delete
find /start/dir -type f -name myfile.txt -print0 | xargs -n1 -0 -I{} rm {}
Build your data science skills into a career

Are you ready to take your data science career to the next step, or break into data science? With Springboard’s Data Science Career Track, you’ll master data science topics, have personalized career guidance, weekly calls with a data science expert, and a job guarantee.


"-print" should be the default action of "find", and "rm -f" is for omitting the confirmation prompt before deleting a write-protected file.

Both things are not vital for the command to work.
Try adding "-print", perhaps your "find" would need it.
Not every find implementation has the "-delete" option, and "xargs -n1" would fork a new instance of rm for each one of the "thousands" of files, which is suboptimal for performance.
Never heard of the need to have a space after rm.

rm also will accept a some filenames list to remove, but if user have a thousands copies of this file - multidelete is not a good solution. To avoid performance issues we can use -P key to run rm processes in parallel or use -n 10 for example.
without space we will get something like 'rmsomefile', not valid 'rm somefile'.
>> without space we will get something like 'rmsomefile', not valid 'rm somefile' <<

Nope. Did you ever run such a command, or is it all theory?
xargs will definitelyadd a space after the command.  The "missing operand" error is probbaly that there were no matching files from the "find" command, so xargs just ran the rm command with no files to delete.   The message does not appear if you give "rm" the "-f" parameter.

If you want to run this more than once, you could run the "find" command to count the number of matching files e.g.

numfiles=$(find /start/dir -name myfile.txt | wc -l)
if [ $numfiles -gt 0 ]
  find /start/dir -name myfile.txt | xargs rm

One other thing to try is to add "-t" to the xargs command - this will print out the command it is about to run, including all of the parameters:

find /start/dir -name myfile.txt | xargs -t rm
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.