python 2.6 support

ext4-inlinedata-calculator.py

#
# This script is a tool that helps you calculate the benefits in occupied size on disk
# of the ext4 inlinedata new feature of the Linux kernel 3.8.0 see http://www.h-online.com/open/features/What-s-new-in-Linux-3-8-1804240.html
# 
# Just run it on your ext4 mountpoints and it will tell give you the trade off for all your files depending on the inode size you # choose.
# 
# To get you current inode size you can do :
# $ tune2fs -l /dev/mapper/isw_ddbeejgcgd_Volume03 | grep Inode
# Inode count:              15040512
# Inodes per group:         8192
# Inode blocks per group:   512
# Inode size:               256
#
# You can set the inode size at creation time with :
# mkfs.ext4 -I inode-size /dev/...
# 

#!/usr/bin/env python
import os
from sys import argv
import string

def find_mount_point(path):
    path = os.path.abspath(path)
    while not os.path.ismount(path):
        path = os.path.dirname(path)
    return path

ext4mountedpaths = []
with open('/etc/mtab') as f:
    for line in f.readlines():
        splitted = string.split(line)
        if splitted[2] == 'ext4':
            ext4mountedpaths.append(splitted[1])


TYPICAL_OCCUPIED_SPACE_FOR_INODE = 124
EXPLORE_INODES_SIZES = (256, 512, 1024, 2048, 4096)

if len(argv) != 2:
    print("Syntax : ext4-inlinedata-calculator.py mountpoint")
    exit(-1)

print('Exploring {0} ...'.format(argv[1]))

allfiles = {}
for root, subFolders, files in os.walk(argv[1]):
    for entry in files:
        try:
            filename = os.path.join(root, entry)
            if find_mount_point(filename) not in ext4mountedpaths:
                continue  # only consider files from a mounted ext4 filesystems

            allfiles[filename] = os.path.getsize(filename) if os.path.isfile(filename) else 0  # count as an empty file everything entry not beeing a real file
            if not (len(allfiles) % 100000):
                print('Read {0} files ...'.format(len(allfiles)))
        except OSError:
            continue  # probably a systemfile

results = {}
for inode_size in EXPLORE_INODES_SIZES:
    results[inode_size] = [0, 0]
total_occupied_space = 0

for _, filesize in allfiles.items():
    total_occupied_space += filesize
    for size in EXPLORE_INODES_SIZES:
        if filesize <= size - TYPICAL_OCCUPIED_SPACE_FOR_INODE:
            results[size][0] += 1
            results[size][1] += size - TYPICAL_OCCUPIED_SPACE_FOR_INODE - filesize
        else:
            results[size][1] += TYPICAL_OCCUPIED_SPACE_FOR_INODE # the full empty space is wasted
if total_occupied_space == 0:
    print("No relevant file found")
    exit(-2)

print("""
== Results ==

Total file size {0:}

Inode size\t\tfiles fit in\t\twasted inode space""".format(total_occupied_space))

total_nb = len(allfiles)
for size in EXPLORE_INODES_SIZES:
    nb, wasted = results[size]
    print('{0}\t\t\t{1:>7.2%}\t\t\t{2:>15.8%}'.format(size, nb / float(total_nb), float(wasted) / float(total_occupied_space)))