Shell script for generic Linux systems. Used to quickly find large files on the file system based on common large file extensions.

find_big_dirs_fast

nlvl=-20; \
printf "MB\tDirectory Name\n"; \
nice -n ${nlvl} \
  find / \
    -type f \
    -regextype egrep \
    -regex ".*\.(gz$|iso$|sql$|bak$|zip$|log$|tar$|csv$|tgz$|mp4$|png$|xz$)" \
    -exec dirname '{}' \; \
| sort -u \
| while read dir; do \
    nice -n ${nlvl} du -m $dir; \
  done 2>&1 \
| sort -nr \
| head -n 20