apache_graphite.py

#!/usr/bin/env python

import sys
import os
import datetime
import socket

MONTH = {
    "Jan":1 ,
    "Feb":2 ,
    "Mar":3 ,
    "Apr":4 ,
    "May":5 ,
    "Jun":6 ,
    "Jul":7 ,
    "Aug":8 ,
    "Sep":9 ,
    "Oct":10,
    "Nov":11,
    "Dec":12,
}


def scale_minute(minute):
    if int(minute[1]) >= 5:
        return minute[0] + '5'
    else:
        return minute[0] + '0'

def parse_apache_date(datestr):
    day,month,yearandtime = datestr.split('/')
    year, hour, minute, second  = yearandtime.split(':')
    return datetime.datetime(int(year),int(MONTH[month]),int(day),int(hour),int(minute))

def update_count_dict(d, k):
    if k in d:
        d[k] += 1
    else:
        d[k] = 1

def parse_apache_log(logfile):
    result = {}
    with open(logfile) as fd:
        for line in fd:
            splited_line = line.split()
            datestr = splited_line[4][1:]
            scaled_date = parse_apache_date(datestr)
            update_count_dict(result, scaled_date.strftime('%s'))
    return result


def send_to_graphite(result):
    key = 'http.count'
    sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    sock.connect(('127.0.0.1', 2003))
    for k,v in result.items():
        sock.send("%s %d %s\n" % (key, v, k))

if __name__ == "__main__":
    data = parse_apache_log(sys.argv[1])
    print data
    send_to_graphite(data)