siege (1) - Linux Manuals


siege is a HTTP/FTP load tester and benchmarking utility.


  siege [options]
  siege [options] <URL>
  siege [options] -g <URL>
  siege [options] -f urls.txt


siege is a multi-threaded HTTP/FTP load tester and benchmarking utility. It supports most of the features detailed in RFCs 2616 (HTTP) and 959 (FTP). Properties can be set at both from the command line and in a configuration file. When the same propertie is set in both locations, the command line takes precedent.

The default configuration file is $HOME/.siege/siege.conf If you don't have a $HOME/.siege directory and a siege.conf and cookies.txt file, siege will generate a new config directory when it runs. You can generate your configu directory with the following command: siege.config


Option Syntax

siege supports long and short options. Short options look like this:
 -c 25

Long options look like this:

Option Values

-V, --version
Displays the siege release version and copyright information.
-h, --help
Prints a help message describing siege's command-line options.
-C, --config
Prints a detailed summary of all the currently configured options, most of which are sent in $HOME/.siege/siege.conf
-v, --vebose
This directive puts siege into verbose mode which is actually a default setting. This command-line option is useful when the config file is set to 'verbose = false' since it will allow you to override that.

By default siege's verbose output is displayed in a color-coded style.
  HTTP 2xx is coded blue
  HTTP 3xx is coded cyan
  HTTP 4xx is coded magenta
  HTTP 5xx is coded red
  HTTP cached is coded black

NOTE: You can turn off color in siege.conf like this: 'color = off'

-q, --quiet
This directive silences siege. It is mostly used for scripting and is often used in conjunction with -g/--get. You can detect the success or failure of the run with its exit code.

  siege --quiet -g
  if [ $? -eq 0 ] ; then
    echo "Success"
    echo "Failure"
-g URL, --get=URL
This option allows you to request a URL and watch the header transaction. There is a corresponding config file directive that allows you to set the request method for these requests: gmethod = HEAD|GET

  $ siege -g ""
  HEAD / HTTP/1.0
  Accept: */*
  User-Agent: Mozilla/5.0 (unknown-x86_64-linux-gnu) Siege/4.0.0-beta5
  Connection: close

  HTTP/1.1 200 OK
  Server: cloudflare-nginx
  Date: Tue, 09 Feb 2016 18:18:41 GMT
  Content-Type: text/html; charset=UTF-8
  Connection: close
  Last-Modified: Wed, 25 Nov 2015 18:46:08 GMT
  Cache-Control: max-age=3, must-revalidate
  Expires: Tue, 09 Feb 2016 18:18:44 GMT
  Vary: Accept-Encoding,Cookie
  CF-RAY: 27219407eeff084a-IAD

NOTE: It's best practice to quote the URL when it's passed to siege from the the command-line.

-c NUM, --concurrent=NUM
This option allows you to set the concurrent number of users. The total number of users is technically limited to your computer's resources.

You should not configure more users than your web server is configured to handle. For example, the default apache configuration is capped at 255 threads. If you run siege with -c 1024, then 769 siege users are left waiting for an apache handler.

For this reason, the default siege configuration is capped at 255 users. You can increase that number inside siege.conf but if you make a mess, then please don't complain to us.

-r NUM, --reps=NUM|once
This option tells each siege user how times it should run. The value should generally be a number greater than zero but it may be the keyword 'once'.

If --reps=3 then each siege user will run three times before it exits. However, if --reps=once, then each user will run through the urls.txt file exactly one time.

For more information about the urls.txt file, see option -f <file>, --file=<file>

-t NUMm, --time=NUMm
This option is similar to --reps but instead of specifying the number of times each user should run, it specifies the amount of time each should run.

The value format is ``NUMm'', where ``NUM'' is an amount of time and the ``m'' modifier is either S, M, or H for seconds, minutes and hours. To run siege for an hour, you could select any one of the following combinations: -t3600S, -t60M, -t1H. The modifier is not case sensitive, but it does require no space between the number and itself.

-d NUM, --delay=NUM
This option instructs siege how long to delay between each page request. The value NUM represents the number of seconds between each one. This number can be a decimal value. In fact the default is half a second (--delay=0.5).

The time between delay requests is NOT applied toward the transaction time. If two 0.1 second transactions have a 2 second delay between them, their average transaction time is run is 0.1 seconds. It is applied toward the total elapsed time. In this scenario, the elapsed time would be 2.2 seconds.

NOTE: when the parser is enabled (see: -p/--parser), there is no delay between the page and its elements, i.e., style sheets, javascripts, etc. The delay is only between page requests.

-b, --benchmark
This directive tells siege to go into benchmark mode. This means there is no delay between iterations.
-i, --internet
This option sets siege into what we call internet mode. It makes requests from the urls.txt file (see: -f <file> / --file=<file>) in random order.
-f FILE, --file=FILE
This option tells siege to work with a list of urls inside a text file. The URLs are listed one per line. Unlike URLs that are passed as a command-line argument, the URLs in this file should not be quoted.

siege's urls.txt parser supports comments and variables.

-R FILE, --rc=FILE
This directive allows you to set an alternative resource file. By default, the siegerc file is $HOME/.siege/siege.conf With this directive, you can override the default and use an alternative file.
-L FILE, --log=FILE
The default log file is $prefix/var/log/siege.log. This directive allows you to specify and alternative file for logging.
-m "string", --mark="string"
This option allows you to log a message to the log file before your stats are written there. It is generally used to identify the proceding run. You could, for example, mark the file with your command-line parameters so it's understood what configuration generated the following data.
-H "header: value", --header="Header: value"
This options allows you to set a custom header in the request. Generally speaking, this request will override an existing header. The Cookie header is a special case. If you set -H ``Cookie: value'' then siege will send that cookie in addition to the other ones.
-A "string", --agent="string"
This option allows you to override the default user-agent with a custom one.

  siege --agent="JoeDog Jr. in da hizzle"

Will set this header:

  User-agent: JoeDog Jr. in da hizzle

Alternatively, you could set the User-agent with the -H/--header option above.

-T "text", --content-type="text"
This is another set header shortcut. You use this option to override the default Content-type request header.


siege supports RFC 1738 URL formats but it takes pains to implement commonly used shortcuts for your convenience. In addition to RFC 1738 formats, siege introduces its own URL format to indicate protocol method.

An RFC 1738 URL looks like this:

A siege URL with a method idicator looks like this:
  <scheme>://<username>:<password>@<hostname>:<port>/<path> POST <query>

You can also post the contents of a file using the redirect character like this:
  <scheme>://<username>:<password>@<hostname>:<port>/<path> POST </home/jeff/haha.txt

Here's two examples with the siege method indicator: POST haha=papa&dada=mama PUT </home/jdfulmer/etc/tests/bbc.jpg

NOTE: If you set URLs with method indicators at the command-line, then you MUST quote the thing or your shell will treat it like three separate arguments. If the URL is in a urls.txt file, then you shouldn't quote it.

As mentioned above, siege goes to great lengths to allow commonly used shortcuts that you're used to from most browser implementations. It treats many parts of the 1738 URL as optional. In this example, the parts in brackets are optional:
  [scheme://] [:port] [/path/file]

When siege receives a host name it builds the URL with default assumptions. becomes


From the section called Option Syntax above we learn that siege can take a URL as an argument. siege -c -r2 will request the JoeDog index page twice. But what if you want to hit large portions of the site? siege will allow you to fill a file with URLs so that it can run through list.

The format for the file is one URL per line: POST homer=simpson&marge=doestoo

The file also supports UNIX-style commenting:
  Comment looks like this POST homer=simpson&marge=doestoo

It supports shell-style variable declaration and references. This is convenient if you want to run the same test on two different tiers or two different shemes:

  $(SCHEME)://$(HOST)/haha/ POST homer=simpson&marge=doestoo

You can tell siege about this file with the -f/--file option:
  siege -c1 -r50 -f /home/jeff/urls.txt


When its run is complete, siege will gather performance data from all its clients and summarize them after the run. (You can also choose to log these numbers). The command-line output is modeled after Lincoln Stein's script:

  Transactions:                   2000 hits
  Availability:                 100.00 %
  Elapsed time:                  58.57 secs
  Data transferred:               5.75 MB
  Response time:                  0.25 secs
  Transaction rate:              34.15 trans/sec
  Throughput:                     0.10 MB/sec
  Concurrency:                    8.45
  Successful transactions:        2000
  Failed transactions:               0
  Longest transaction:            4.62
  Shortest transaction:           0.00

      This number represents the total number of HTTP requests. In this
      example, we ran 25 simulated users [-c25] and each ran ten times
      [-r10]. Twenty-five times ten equals 250 so why is the transaction 
      total 2000? That's because siege counts every request. This run 
      included a META redirect, a 301 redirect and the page it requested 
      contained several elements that were also downloaded.

      This is the percentage of socket connections successfully handled 
      by the server. It is the result of socket failures (including 
      timeouts) divided by the sum of all connection attempts. This 
      number does not include 400 and 500 level server errors which are
      recorded in "Failed transactions" described below.  

  Elapsed time
      The duration of the entire siege test. This is measured from the 
      time the user invokes siege until the last simulated user 
      completes its transactions. Shown above, the test took 14.67 
      seconds to complete.

  Data transferred
      The sum of data transferred to every siege simulated user. It 
      includes the header information as well as content. Because it
      includes header information, the number reported by siege will 
      be larger then the number reported by the server. In internet 
      mode, which hits random URLs in a configuration file, this 
      number is expected to vary from run to run.

  Response time
      The average time it took to respond to each simulated user's requests.

  Transaction rate
      The average number of transactions the server was able to handle
      per second, in a nutshell: it is the count of all transactions 
      divided by elapsed time.

      The average number of bytes transferred every second from the 
      server to all the simulated users.

      This is the average number of simultaneous connections. The metric
      is calculated like this: the sum of all transaction times divided
      by elapsed time (how long siege ran) 

  Successful transactions
      The number of times the server responded with a return code < 400.

  Failed transactions
      The number of times the socket transactions failed which includes 
      socket timeouts.

  Longest transaction
      The greatest amount of time that any single transaction took, out 
      of all transactions.

  Shortest transaction
      The smallest amount of time that any single transaction took, out
      of all transactions.


Jeffrey Fulmer, et al. <jeff [at]> is the primary author of siege. Numerous people throughout the globe also contributed to this program. Their contributions are noted in the source code ChangeLog


Copyright by Jeffrey Fulmer, et al. <jeff [at]>

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.


The most recent released version of siege is available by HTTP download: