Find large files on Linux.

UPDATE: Reader Luc Pionchon points out that sort often supports a -h parameter that sorts by "human" numbers, hence:

$ du -h * | sort -h | tail

is probably a better alternative than any of the following (for the systems that support it).

du -h * | grep "^[0-9.]*M" | sort -n

This finds files at least 1 MB in size and then sorts them by size. Change M to G for files at least 1 GB in size.

(Caveat: files 1 GB or larger will be missed by the MB version. You can use:

du -h * | egrep "^[0-9.]*(M|G)"

to get both, but then the sort -n doesn't work quite the way we'd like.)

Of course, you could use du without the -h to get file sizes by the default block size rather than the human-readable 12.4M or 16K, etc.