Centos – impossible creating a ext4 fs with block size of 1024

Here are the answers that are true for ext2/ext3/ext4. If they are true for other file systems depends on their implementation.

1. user48838 answered this one correctly. More files consume more meta data. They are allocated in 4k chunks or in any other size defined at creation time of the file system
2. Yes it is a feature/problem of the real file system
3. In an ext3 file system this is not possible. Only by recreating the (empty) directory
4. The source code is around here and in related files

But you have luck. When you recreate the same amount of files you already deleted, the directory size will stay the same. Only when you add more files it will increase.

The following one-liner will give you a listing of how many files are in each directory, and sort by the top ten. It will run recursively from your current working directory, so I don't suggest you run this from / unless you have absolutely no clue where the large directories may be.

find . -type f | awk '{dir=gensub(/(.+\/).+/,"\\1","g (file://1%22,%22g/)"); dir_list[dir]++} END {for (d in dir_list) printf "%s %s\n",dir_list[d],d}d' | sort -nr |head 

Output will be similar to the following:

[user@localhost ~]# find . -type f | awk '{dir=gensub(/(.+\/).+/,"\\1","g (file://1%22,%22g/)"); dir_list[dir]++} END {for (d in dir_list) printf "%s %s\n",dir_list[d],d}d' | sort -nr | head
2048 ./test19/
2048 ./test18/
2048 ./test17/
2048 ./test16/
2048 ./test15/
2048 ./test14/
2048 ./test13/
2048 ./test12/
2048 ./test11/
2048 ./test10/

If you're a bit wary about running such a one-line just search for all directories which themselves have a size of over 50k or so. Again find will be your friend here:

find ./ -type d -size +50k

If you have multiple mount points, a df -i will help you narrow down which mount is running out of, or has run out of, inodes.