Hi,

I will copy files from /test to /test2

with cron and use $date to create names. I created a bash for that.

Every week it will copy one file from /test to /test2

But I want after few weeks to delete old files and keep only last 6 ones, because of used spaces. I have 1TB and files are 100GB.

Is there any way to do it autoamticly by cron job?

How often are you creating the files in /test2? Can you say they're being created regularly?

Let's say that within 14 days you create six files. Anything older than than can be deleted, so simple find . -mtime +14 | xargs rm should work no problem.

For example I want to delete all file in a folder /etc/test created before 40 days or more.

I dont understand really your example. Can you use code tags and create it again please?

Something like that?

Kind of:

I just found another example. But which is the different?

Both do the same?

-exec function of find executes rm on each file. With a lot of files, xargs can be noticeably more efficient. I'd use -0 switch if spaces may appear in file path/name.

As for the -type f, that's just personal choice. rm won't delete folders without -r, but type -f[ile] just makes sure folder names aren't passed to rm anyway.

I'd change your example slightly:

Thank you rmacd,

New un LQ but I see you have a lot of experience in Linux.

This is not quite true, how the -exec option behaves is determined by the ending character of the command. Usually you will use the ; ending character (which due to its function as a special character in the shell has to be escaped to \;), but you can also use the + ending character, which will behave similarly to xargs.

So you have several options to achieve the task, the most common would be:

1. If it is expected that the find command will find only a few files (which is the case for the OP) you can useorThis will, as rmacd already stated, invoke the deletion on any found file separately, which will cause big overhead when many files are found.

2. If it is expected that a large number of files will be found it is better to use eitherorIn both cases the found files will be assembled to sets of files and then the command is launched on a whole set of filenames(in the first case by the xargs command, in the second find will do the work). This will speed up the execution when large numbers of files are found.

2 members found this post helpful.

I tried your examples but I have got this errors:

Any idea?

Tobi,
Only this command works for me

Other ones that end with RM doesnot work.

And is there any way to add some parameter, to see what happend after command is executed?
With rsync I can add -v and I can see that happend, but with find, I dont see -v exist.

After I execute
there is nothing in the log file. It is empty.

Any idea?