yahi

yahi er en alsidig log parser levere standard udsugningsanlæg til apache / lighttpd.
brug Command line
Eksempel på data parsede med yahi: http://wwwstat.julbox.fr/
Enkleste brug er:
speed_shoot -g / usr / local / data / GeoIP / var / www / apache / adgang * log
det vil vende tilbage en JSON i form:
{
& Nbsp; "by_date": {
& Nbsp; "2012/05/03": 11
& Nbsp;},
& Nbsp; "total_line": 11,
& Nbsp; "ip_by_url": {
& Nbsp; "/ favicon.ico": {
& Nbsp; "192.168.0.254": 2,
& nbsp; "192.168.0.35": 2
& Nbsp;},
& Nbsp; "/": {
& Nbsp; "74.125.18.162": 1,
& Nbsp; "192.168.0.254": 1,
& nbsp; "192.168.0.35": 5
& Nbsp;}
& Nbsp;},
& Nbsp; "by_status": {
& Nbsp; "200": 7,
& Nbsp; "404": 4
& Nbsp;},
& Nbsp; "by_dist": {
& Nbsp; "ukendt": 11
& Nbsp;},
& Nbsp; "bytes_by_ip": {
& Nbsp; "74.125.18.162": 151,
& Nbsp; "192.168.0.254": 489,
& Nbsp; "192.168.0.35": 1093
& Nbsp;},
& Nbsp; "by_url": {
& Nbsp; "/ favicon.ico": 4,
& Nbsp; "/": 7
& Nbsp;},
& nbsp; "by_os": {
& Nbsp; "ukendt": 11
& Nbsp;},
& Nbsp; "week_browser": {
& Nbsp; "3": {
& Nbsp; "ukendt": 11
& Nbsp;}
& Nbsp;},
& Nbsp; "by_referer": {
& Nbsp; "-": 11
& Nbsp;},
& Nbsp; "by_browser": {
& Nbsp; "ukendt": 11
& Nbsp;},
& Nbsp; "by_ip": {
& Nbsp; "74.125.18.162": 1,
& Nbsp; "192.168.0.254": 3,
& nbsp; "192.168.0.35": 7
& Nbsp;},
& Nbsp; "by_agent": {
& Nbsp; "Mozilla / 5.0 (X11, Ubuntu, Linux x86_64; rv: 12.0) Gecko / 20100101 Firefox / 12.0, gzip (GFE) (via translate.google.com)": 1,
& Nbsp; "Mozilla / 5.0 (X11, Ubuntu, Linux x86_64; rv: 12.0) Gecko / 20100101 Firefox / 12.0": 10
& Nbsp;},
& Nbsp; "by_hour": {
& Nbsp; "9": 3,
& Nbsp; "10": 4,
& Nbsp; "11": 1,
& Nbsp; "12": 3
& Nbsp;},
& Nbsp; "by_country": {
& Nbsp; "": 10,
& Nbsp; "USA": 1
& Nbsp;}
}
Hvis du bruger:
speed_shoot -f csv -g / usr / local / data / GeoIP / var / www / apache / adgang * log
Dit resultat er:
by_date, 2012-5-3,11
total_line, 11
ip_by_url, / favicon.ico, 192.168.0.254,2
ip_by_url, / favicon.ico, 192.168.0.35,2
ip_by_url, /, 74.125.18.162,1
ip_by_url, /, 192.168.0.254,1
ip_by_url, /, 192.168.0.35,5
by_status, 200,7
by_status, 404,4
by_dist, ukendt, 11
bytes_by_ip, 74.125.18.162,151
bytes_by_ip, 192.168.0.254,489
bytes_by_ip, 192.168.0.35,1093
by_url, / favicon.ico, 4
by_url, /, 7
by_os, ukendt, 11
week_browser, 3, ukendt, 11
by_referer, -, 11
by_browser, ukendt, 11
by_ip, 74.125.18.162,1
by_ip, 192.168.0.254,3
by_ip, 192.168.0.35,7
by_agent, "Mozilla / 5.0 (X11, Ubuntu, Linux x86_64; rv: 12.0) Gecko / 20100101 Firefox / 12.0, gzip (GFE) (via translate.google.com)", 1
by_agent, Mozilla / 5.0 (X11, Ubuntu, Linux x86_64; rv: 12.0) Gecko / 20100101 Firefox / 12.0,10
by_hour, 9,3
by_hour, 10,4
by_hour, 11,1
by_hour, 12,3
by_country ,, 10
by_country, USA, 1
Jeg gætter godt, det virker ikke, fordi du først nødt til at hente GeoIP datafil:
wget -O "http://www.maxmind.com/download/geoip/database/GeoLiteCountry/GeoIP.dat.gz" | zcat> /usr/local/data/GeoIP.dat
Selvfølgelig er dette den geoLite databasen jeg ikke indeholde data i pakken siden GeoIP skal opdateres ofte at blive præcis.
Standard sti for GeoIP er data / GeoIP.dat
Anvendelse som script
speed shoot er i virkeligheden en skabelon af hvordan man bruger yahi som et modul:
#! / Usr / bin / env python
fra archery.bow import Hankyu som _dict
fra yahi import hak, skyde
fra datetime import datetime
context = hak ()
date_formater = lambda dt: "% s-% s-% s"% (dt.year, dt.month, dt.day)
context.output (
& Nbsp; skyde (
& Nbsp; kontekst,
& Nbsp; lambda data: _dict ({
& Nbsp; "by_country«: _dict ({data ['_ land']: 1}),
& Nbsp; "by_date«: _dict ({date_formater (data ['_ datetime']): 1}),
& Nbsp; "by_hour«: _dict ({. Data ['_ datetime'] hour: 1}),
& Nbsp; "by_os«: _dict ({data ['_ os_name']: 1}),
& Nbsp; "by_dist«: _dict ({data ['_ dist_name']: 1}),
& Nbsp; "by_browser«: _dict ({data ['_ BROWSER_NAME']: 1}),
& Nbsp; "by_ip«: _dict ({data ['ip']: 1}),
& nbsp; "by_status«: _dict ({data ['status']: 1}),
& Nbsp; "by_url«: _dict ({data ['Uri']: 1}),
& Nbsp; "by_agent«: _dict ({data ['agent']: 1}),
& Nbsp; "by_referer«: _dict ({data ['referer']: 1}),
& Nbsp; "ip_by_url«: _dict ({data ['Uri']: _dict ({data ['ip']: 1})}),
& Nbsp; "bytes_by_ip«: _dict ({data ['ip']: int (data ['bytes'])}),
& Nbsp; "week_browser«:. _dict ({Data ['_ datetime'] ugedag ():
& Nbsp; _dict ({data ["_ BROWSER_NAME"]: 1})}),
& Nbsp; "total_line«: 1,
& Nbsp;}),
& Nbsp;)
)
Installation
let som:
pip installere yahi
eller:
easy_install yahi
Anbefalte forbrug
- For grundlæggende log sammenlægning, jeg kan recommand at bruge kommandolinjen;
- For hver enkelt optagelse målinger jeg anbefale en interaktiv konsol (bpython eller ipython);
. - For bestemte målinger eller kunstfærdige filtre Jeg recommand at bruge API

Krav :

• Python