Browsermob::Proxy - Perl client for the proxies created by the Browsermob server


version 0.17



    my $proxy = Browsermob::Proxy->new(
        server_port => 9090
        # port => 9092

    print $proxy->port;
    # create network traffic across your port
    $proxy->har; # returns a HAR as a hashref, converted from JSON

with Browsermob::Server:

    my $server = Browsermob::Server->new(
        server_port => 9090
    $server->start; # ignore if your server is already running

    my $proxy = $server->create_proxy;
    $proxy->new_har('proxy from server!');



    BrowserMob proxy is based on technology developed in the Selenium open source project and a commercial load testing and monitoring service originally called BrowserMob and now part of Neustar.

    It can capture performance data for web apps (via the HAR format), as well as manipulate browser behavior and traffic, such as whitelisting and blacklisting content, simulating network traffic and latency, and rewriting HTTP requests and responses.

This module is a Perl client interface to interact with the server and its proxies. It uses Net::HTTP::Spore. You can use Browsermob::Server to manage the server itself in addition to using this module to handle the proxies.


We depend on Net::HTTP::Spore to set up our communication with the Browsermob server. Unfortunately, there hasn't been a recent release and due to breaking changes in new versions of its dependencies, you might run in to problems installing its current CPAN version v0.06. And, thus installing this module may be difficult.

We're using a fork of Net::HTTP::Spore that is kept slightly ahead of master with the bug fixes merged in; installation via App::cpanminus looks like:

    cpanm git://



Optional: specify where the proxy server is; defaults to

    my $proxy = Browsermob::Proxy->new(server_addr => '');


Optional: Indicate at what port we should expect a Browsermob Server to be running; defaults to 8080

    my $proxy = Browsermob::Proxy->new(server_port => 8080);


Optional: When instantiating a proxy, you can choose the proxy port on your own, or let the server automatically assign you an unused port.

    my $proxy = Browsermob::Proxy->new(port => 9091);


Set Net::HTTP::Spore's trace option; defaults to 0; set it to 1 to see headers and 2 to see headers and responses. This can only be set during construction; changing it afterwards will have no impact.

    my $proxy = Browsermob::Proxy->new( trace => 2 );



After creating a proxy, new_har creates a new HAR attached to the proxy and returns the HAR content if there was a previous one. If no argument is passed, the initial page ref will be "Page 1"; you can also pass a string to choose your own initial page ref.


This convenience method is just a helper around the actual endpoint method /create_new_har; it uses the defaults of not capturing headers, request/response bodies, or binary content. If you'd like to capture those items, you can use create_new_har as follows:

        payload => {
            initialPageRef => 'payload is optional'
        captureHeaders => 'true',
        captureContent => 'true',
        captureBinaryContent => 'true'


After creating a proxy and initiating a new_har, you can retrieve the contents of the current HAR with this method. It returns a hashref HAR, and may in the future return an isntance of Archive::HAR.

    my $har = $proxy->har;
    print Dumper $har->{log}->{entries}->[0];


Generate the proper capabilities for use in the constructor of a new Selenium::Remote::Driver object.

    my $proxy = Browsermob::Proxy->new;
    my $driver = Selenium::Remote::Driver->new(
        browser_name => 'chrome',
        proxy        => $proxy->selenium_proxy
    print Dumper $proxy->har;

N.B.: selenium_proxy will AUTOMATICALLY call "new_har" for you initiating an unnamed har, unless you pass it something truthy.

    my $proxy = Browsermob::Proxy->new;
    my $driver = Selenium::Remote::Driver->new(
        browser_name => 'chrome',
        proxy        => $proxy->selenium_proxy(1)
    # later
    print Dumper $proxy->har;


Generate a hash with the proper keys and values that for use in setting preferences for a Selenium::Remote::Driver::Firefox::Profile. This method returns a hashref; dereference it when you pass it to "set_preference" in Selenium::Remote::Driver::Firefox::Profile:

    my $profile = Selenium::Remote::Driver::Firefox::Profile->new;

    my $firefox_pref = $proxy->firefox_proxy;
    $profile->set_preference( %{ $firefox_pref } );

    my $driver = Selenium::Remote::Driver->new_from_caps(
        desired_capabilities => {
            browserName => 'Firefox',
            firefox_profile => $profile->_encode

N.B.: firefox_proxy will AUTOMATICALLY call "new_har" for you initiating an unnamed har, unless you pass it something truthy.


Generate the proper arguments for the proxy method of LWP::UserAgent. By default, ua_proxy will initiate a new har for you automatically, the same as "selenium_proxy" does. If you want to initialize the har yourself, pass in something truthy.

    my $proxy = Browsermob::Proxy->new;
    my $ua = LWP::UserAgent->new;


Export to %ENV the properties of this proxy's port. This can be used in tandem with <LWP::UserAgent/env_proxy>. This will set the appropriate environment variables, and then your $ua will pick it up when its env_proxy method is invoked aftewards. As usual, this will create a new HAR unless you deliberately inhibit it.


In particular, we set http_proxy, https_proxy, and ssl_proxy to the appropriate server and port by defining them as keys in %ENV.


Set up automatic Basic authentication for a specified domain. Accepts as input a HASHREF with the keys domain, username, and password. For example,

        domain => '',
        username => 'username',
        password => 'password'

set_request_header ( $header, $value )

Takes two STRINGs as arguments. (Unhelpfully) returns a Net::HTTP::Spore::Response. With this method, we will remove the specified $header from every request the proxy sees, and replace it with the $header $value pair that you pass in.

    $proxy->set_request_header( 'User-Agent', 'superwoman' );

Under the covers, we are using "filter_request" with a Javascript Rhino payload.

set_timeout ( $timeoutType => $milliseconds )

Set different time outs on the instantiated proxy. You can set multiple timeouts at once, if you like.

        requestTimeout => 5000,
        readTimeout => 6000
  • requestTimeout

    Request timeout in milliseconds. A timeout value of -1 is interpreted as infinite timeout. It equals -1 by default.

  • readTimeout

    Read timeout is the timeout for waiting for data or, put differently, a maximum period inactivity between two consecutive data packets. A timeout value of zero is interpreted as an infinite timeout. It equals 60000 by default.

  • connectionTimeout

    Determines the timeout in milliseconds until a connection is established. A timeout value of zero is interpreted as an infinite timeout. It eqauls 60000 by default.

  • dnsCacheTimeout

    Sets the maximum length of time that records will be stored in this Cache. A nonpositive value disables this feature (that is, sets no limit). It equals 0 by default.


Delete the proxy off of the server, shutting down the port. Although we do try to do this in our DEMOLISH method, we can't do anything if the $proxy object is kept around during global destruction. If you're noticing that your BMP server has leftover proxies, you should start either explicitly undefing the `$proxy` object or invoking this method.

    # calls ->delete_proxy in our DEMOLISH method, explicitly not
    # during global destruction!
    undef $proxy;

    # manually delete the proxy from the BMP server

After deleting the proxy, invoking any other method will probably lead to a die from inside the Net::HTTP::Spore module somewhere.


Please see those modules/websites for more information related to this module.


Please report any bugs or feature requests on the bugtracker website

When submitting a bug or request, please include a test-file or a patch to an existing test-file that illustrates the bug or desired feature.


Daniel Gempesaw <>


This software is copyright (c) 2014 by Daniel Gempesaw.

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.