Que és Apache Benchmark?
Tens una web personal a un servidor web i vols saber quin rendiment té? Doncs bé, a Linux hi ha una utilitat que és diu Apache Benchmarking (ab) que serveix per realitzar proves de càrrega del nostre servidor web, o sigui que és una utilitat que esta dissenyada per saber com es comporta la configuració actual de nostre servidor web. Ara bé no és tracta de simplement de saber el temps de càrrega d’una pàgina ja que hi ha retards en la xarxa i això pot fer que els resultats obtinguts en un moment variin considerablement als d’un altre. El que si que pot detectar aquesta utilitat és el temps de càrrega màxim per a un número determinat d’usuaris que intenten accedir a la teva pàgina.També pot ser interessant mirar el temps addicional que pot trigar en carregar la mateixa pàgina si per exemple es duplica el número de visitants.
Les proves que es facin amb ab s’han de repetir vàries vegades per a evitar desviacions i així quedar-nos amb el millor resultat.
Els possibles paràmetres que accepta ab són:
ab [ -A auth-username:password ] [ -c concurrency ] [ -C cookie-name=value ] [ -d ] [ -e csv-file ] [ -g gnuplot-file ] [ -h ] [ -H custom-header ] [ -i ] [ -k ] [ -n requests ] [ -p POST-file ] [ -P proxy-auth-username:password ] [ -q ] [ -s ] [ -S ] [ -t timelimit ] [ -T content-type ] [ -v verbosity] [ -V ] [ -w ] [ -x <table>-attributes ] [ -X proxy[:port] ] [ -y <tr>-attributes ] [ -z <td>-attributes ] [http://]hostname[:port]/path
-A auth-username:password Supply BASIC Authentication credentials to the server. The username and password are separated by a single : and sent on the wire base64 encoded. The string is sent regardless of whether the server needs it (i.e., has sent an 401 authentication needed). -c concurrency Number of multiple requests to perform at a time. Default is one request at a time. -C cookie-name=value Add a Cookie: line to the request. The argument is typically in the form of a name=value pair. This field is repeatable. -d Do not display the "percentage served within XX [ms] table". (legacy support). -e csv-file Write a Comma separated value (CSV) file which contains for each percentage (from 1% to 100%) the time (in milliseconds) it took to serve that percentage of the requests. This is usually more use- ful than the ’gnuplot’ file; as the results are already ’binned’. -g gnuplot-file Write all measured values out as a ’gnuplot’ or TSV (Tab separate values) file. This file can eas- ily be imported into packages like Gnuplot, IDL, Mathematica, Igor or even Excel. The labels are on the first line of the file. -h Display usage information. -H custom-header Append extra headers to the request. The argument is typically in the form of a valid header line, containing a colon-separated field-value pair (i.e., "Accept-Encoding: zip/zop;8bit"). -i Do HEAD requests instead of GET. -k Enable the HTTP KeepAlive feature, i.e., perform multiple requests within one HTTP session. Default is no KeepAlive. -n requests Number of requests to perform for the benchmarking session. The default is to just perform a sin- gle request which usually leads to non-representative benchmarking results. -p POST-file File containing data to POST. -P proxy-auth-username:password Supply BASIC Authentication credentials to a proxy en-route. The username and password are sepa- rated by a single : and sent on the wire base64 encoded. The string is sent regardless of whether the proxy needs it (i.e., has sent an 407 proxy authentication needed). -q When processing more than 150 requests, ab outputs a progress count on stderr every 10% or 100 requests or so. The -q flag will suppress these messages. -s When compiled in (ab -h will show you) use the SSL protected https rather than the http protocol. This feature is experimental and very rudimentary. You probably do not want to use it. -S Do not display the median and standard deviation values, nor display the warning/error messages when the average and median are more than one or two times the standard deviation apart. And default to the min/avg/max values. (legacy support). -t timelimit Maximum number of seconds to spend for benchmarking. This implies a -n 50000 internally. Use this to benchmark the server within a fixed total amount of time. Per default there is no timelimit. -T content-type Content-type header to use for POST data. -v verbosity Set verbosity level - 4 and above prints information on headers, 3 and above prints response codes (404, 200, etc.), 2 and above prints warnings and info.
-V Display version number and exit. -w Print out results in HTML tables. Default table is two columns wide, with a white background. -x <table>-attributes String to use as attributes for <table>. Attributes are inserted <table here >. -X proxy[:port] Use a proxy server for the requests. -y <tr>-attributes String to use as attributes for <tr>. -z <td>-attributes String to use as attributes for <td>.
Per a realitzar les proves amb aquesta utilitat és important que la prova es faci des d’un altre màquina que no formi part de la xarxa del servidor que volem provar ja que sinó els temps que obtindríem serien evidentment més petits dels esperats.
En el meu cas he fet la prova des el meu ordinador d’escriptori fins al servidor on esta allotjat aquest blog. Concretament sobre la següent URL http://prova.jjberdullas.com. Que serveix una pàgina tant simple com aquesta:
<html> <head> <title>prova</title> </head> <body> <?php echo "aixo es una prova"; ?> </body> </html>
Quan es crida la comanda ab s’ha d’afegir una barra final a la URL sobre la qual es vol fer la prova.
Sobre les quals he fet 2 proves diferents:
- 100 consultes amb 20 usuaris simultanis
[usuari@nom_servidor ~]$ ab -n 1000 -c 20 http://prova.jjberdullas.com/ Server Software: Apache/X.X.X Server Hostname: prova.jjberdullas.com Server Port: 80 Document Path: / Document Length: 84 bytes Concurrency Level: 20 Time taken for tests: 12.860595 seconds Complete requests: 1000 Failed requests: 0 Write errors: 0 Total transferred: 276276 bytes HTML transferred: 84084 bytes Requests per second: 77.76 [#/sec] (mean) Time per request: 257.212 [ms] (mean) Time per request: 12.861 [ms] (mean, across all concurrent requests) Transfer rate: 20.92 [Kbytes/sec] received Connection Times (ms) min mean[+/-sd] median max Connect: 42 115 35.2 127 285 Processing: 46 139 32.5 141 290 Waiting: 45 138 32.6 140 289 Total: 89 254 43.1 247 532
- 2000 consultes amb 20 usuaris simultanis
[usuari@nom_servidor ~]$ ab -n 2000 -c 20 http://prova.jjberdullas.com/ Server Software: Apache/X.X.X Server Hostname: prova.jjberdullas.com Server Port: 80 Document Path: / Document Length: 84 bytes Concurrency Level: 20 Time taken for tests: 25.16090 seconds Complete requests: 2000 Failed requests: 0 Write errors: 0 Total transferred: 552000 bytes HTML transferred: 168000 bytes Requests per second: 79.95 [#/sec] (mean) Time per request: 250.161 [ms] (mean) Time per request: 12.508 [ms] (mean, across all concurrent requests) Transfer rate: 21.55 [Kbytes/sec] received Connection Times (ms) min mean[+/-sd] median max Connect: 42 119 21.9 132 147 Processing: 43 129 22.2 134 234 Waiting: 42 127 22.2 133 234 Total: 85 248 22.2 246 328 Percentage of the requests served within a certain time (ms) 50% 246 66% 250 75% 256 80% 274 90% 279 95% 282 98% 285 99% 287 100% 328 (longest request)
De lo cual es pot deduir que amb el mateix número d’usuaris concurrents i el doble de peticions, es duplica el temps per realitzar la prova però amb la configuració actual del nostre servidor, es manté el número de peticions per segon que són ateses (aproximadament 75), temps per petició i el temps per petició sobre totes les peticions concurrents.