surcharge 0.9.1

Surcharge is a tool for benchmarking your web server

Issues

• ProgressBar issues

• DNS resolution is wrong

Introduction

Surcharge is a tool for benchmarking your web server like apache benchmark. Surcharge uses the gevent networking library. Using the greenlets allow to spawn many concurrent requests with little memory. HTTP requests are made with requests library.

Requirements

This code has been run on Python 2.7

# install by the setup
greenlet==0.4.0
gevent==0.13.6
requests==1.2.3
progressbar==2.3

Installation

$ pip install surcharge #and enjoy

Tests

Example command

# simple call
$ surcharge http://google.com --numbers 10 --concurrency 5

# standard stdout
Server: gws

URL: http://173.194.67.138:80

Concurrency level: 5

Options: {'cookies': {}, 'timeout': 2.0}


100% |############################|


Number process requests: 10
Time taken for tests: 0.57
Complete requests: 10
Failed requests: 0
Faster request: 0.045
Slower request: 0.059
Time per request (only success): 0.051
Request per second: 98.57

# call with multiple cookies
$ surcharge http://httpbin.org/cookies --cookies "{'ck':1, 'cook':value}"

# call with HTTP Basic Auth
$ surcharge https://secure.test.com --auth "user:password"

# bench during 10 seconds
$ surcharge http://google.com --concurrency 10 --duration 10

Example API

 # see the constructor or the surcharge/cli.py for more details
 >>> from surcharge.core import Surcharger
 >>> surcharge = Surcharger(url='http://google.com')
 >>> surcharge()

 >>> surcharge.result
 defaultdict(<type 'list'>, {200: [0.06690406799316406]})

 # compute simple stat
 >>> from surcharge.core import SurchargerStats
 >>> surcharge_stats = SurchargerStats(surcharge)
 >>> surcharge_stats()

 >>> surcharge_stats.stats
 {'RPS': 14.20353538774128,
'exec_time': 0.07088184356689453,
'max': 0.0704050064086914,
'min': 0.0704050064086914,
'moy': 0.0704050064086914,
'requests_process': 0.0704050064086914,
'total': 1,
'total_failed': 0,
'total_success': 1}

 # By default, stdout is used to display the stats
 # You can override the SurchargerStats.send method and make what you want with the stats

 # example with a *request" option (auth)
 >>> surcharge = Surcharger(url='http://google.com', **{'auth': ('user', 'pass')})

Usage

Usage:
    surcharge <url>
        [--method=<method>]
        [--concurrency=<clients>]
        [--numbers=<requests> | --duration=<seconds>]
        [--timeout=<seconds>]
        [--cookies=<cookies>]

Options:
    -h --help                           Show this screen.
    -v --version                        Show version.
    -m --method=<method>                HTTP method [default: GET].
    -c --concurrency=<clients>          Number of multiple requests to perform at a time [default: 1].
    -n --numbers=<requests>             Number of requests to perform for the benchmarking session [default: 1].
    -D --duration=<seconds>             Duration in seconds. Override the --numbers option [default: 0]
    -T --timeout=<seconds>              You can tell requests to stop waiting for a response after a given number of seconds [default: 2].
    -C --cookies=<cookies>              Send your own cookies. [default: {}]

Logs

You can retrieve some informations in the /tmp/surcharge_activity.log log file.

Deprecated - need refactoring

(since 0.8) Overflow is a module of Surcharge. It’s allows to launch several benchmarks dynamically through the network. Overflow uses the zeroMQ library.

License

This project is lecensed under the MIT license, a copy of which can be found in the LICENSE file.