I have a directory that keeps filling up and i need to clean it up. So i want to find large files i can possibly delete. I need a find command or shell script that will find large files over 500m and sort them with the largest at the top.

Im using one that isnt working very well:

#!/bin/bash
# if nothing is passed to the script, show usage and exit
[[ -n "$1" ]] || { echo "Usage: findlarge [PATHNAME]"; exit 0 ; }
# simple using find, $1 is the first variable passed to the script
find $1 -type f -size +520000k -exec ls -lh {} \; | awk '{ print $9 ": " $5 }'

Thanks

Hi,

You don't mention which problems you are experiencing. And I have to assume it has to do with files with spaces in them (your script otherwise works).

Maybe this will help, replace your find statement (including the pipe and awk) with:

find "$1" -type f -size +520000k -exec du -h {} \;

Instead of using ls and pipe all found to awk, it uses du (-h for human readability). The layout is a bit different then your script. Instead of
filename : 1Gb
it shows
1G filename

Hope this helps.

How about trying

find $1 -type f -size +520000k -exec ls -l {} \; | awk '{ print $5 " " $9 }' | sort -nr

this will do a reverse numeric sort

BTW: don't forget that space is freed only when no more processes are accessing the file.

Cleaning up by a script as an automated process is quite like betting on good luck. Why not find out where those unneeded large files do come from and what user or application is responsible for them? Would make more sense to me.

I would use 'du -h' and sort the results by file size with awk or something similar.

Linux Archive