Sun Java System Web Server 7.0 Update 4 Administrator's Guide

Configuring Reverse Proxy in Web Server 7.0

The Sun Java System Web Server 7.0 integrates the reverse proxy functionality within the core server.

When web server is configured with reverse proxy functionality, it acts as a proxy for one or more backend servers and serves as a single point of access or gateway in a server farm. In a reverse proxy setup, the web server forwards the HTTP request it received from the browser client to the appropriate backend server. The HTML response from the backend server is sent back to the browser through the web server. Thus, the web server with reverse proxy hides the existence of backend servers.

Figure 4–4 Reverse Proxy Setup

Reverse proxy setup

Web Server 7.0 with reverse proxy functionality acts as a simple software load balancer with the added ability to forward the sticky requests back to the same backend server.

Web server with reverse proxy can serve static content like gif and html files from its internal cache. At the same time, it functions as a load balancer and processes request for dynamic content like jsp, servlet or php files to the backend server. When web server is deployed in this configuration, disabling the Java web container will significantly reduce the memory footprint of the server. For information about disabling Java web container, see Tuning Web Container Within Web Server 7.0 in Sun Java System Web Server 7.0 Update 4 Performance Tuning, Sizing, and Scaling Guide. See CLI Reference, disable-java(1).

The advantages of reverse proxy within Web Server 7.0 are:

Note –

In a typical deployment, one or more reverse proxies will be deployed between the browsers and the backend servers.

Configuring Reverse Proxy for Load-balancing

Web Server 7.0 provides a sophisticated built-in load balancer, the reverse proxy, which distributes load or request from the client to several backend servers.

Web Server provides GUI and CLI support for configuring the reverse proxy.

ProcedureConfiguring Reverse Proxy Using Administration Console

  1. Install Web Server on the node that you want to use for configuring reverse proxy.

  2. Create a configuration. For example, rp.

  3. Using the Administration Console, select Configurations > Virtual Servers > Content Handling > Reverse Proxy tab. Click New.

  4. Specify values for the following parameters:

    • URI — The reverse proxy URI

    • Server URL — Comma separated server URLs of all the machines in the cluster separated by comma. If multiple values are given, the server will distribute load among the specified servers.

      The format for entering the server URL is hostname:portnumber. For example, http://<content server-hostname>:port

  5. Click the OK button.

  6. Click the Deployment Pending link in the top right of the screen to deploy the modified configuration and to apply changes to the configuration.

  7. Click the Deploy button.

    Deployment successful message appears.

  8. Start all instances of this modified configuration.

    This completes configuring the reverse proxy for load balancing HTTP requests.

    Note –

    To configure a reverse proxy in a cluster environment, issue a wildcard server certificate or the alternate subject names that can be set to the actual origin server host names. The other option of specifying the original server's host names in the subject name field limits the size of the cluster, leading the cluster to fail if another node is added to the cluster.

    A wildcard server certificate can be created using the administration interfaces. After creating the server certificate use certutil to get the base64 encoded version of the certificate and install it as a trusted CA certificate on the load balancer configuration.

    Type the following command to generate the base64 encoded certificate bash$./certutil -L -a -d instancedir/config. Copy the output of the command and paste it in the install certificate wizard.

ProcedureConfiguring Reverse Proxy Using CLI

Perform the following steps to configure reverse proxy in CLI mode. You will create a configuration config1 and an instance rp as reverse proxy.

  1. Start the Administration Server:

    $ <install-dir>/admin-server/bin/startserv

  2. Invoke the CLI shell:

    <install-dir>/admin-server/bin/wadm -user <username>

    You can see the wadm shell

  3. Create config1:

    wadm>create-config --http-port=8080 --server-name=config1 --server-user=root config1

  4. Create an instance for the config1 configuration:

    wadm>create-instance --config=config1 <host-name>

  5. Add the web application on the created configuration:

    wadm>add-webapp --config=config1 -vs=config1 --uri/test <warfile>

  6. Deploy the web application.

    wadm>deploy-config --user=admin --password-file=admin.pwd --host=serverhost --port=8989 config1

  7. Create a rp configuration:

    wadm>create-config --http-port=8081 --server-name=rp --server-user=root rp

  8. Enable the rp configuration to reverse proxy using the following command:

    wadm> create-reverse-proxy --user=admin --password-file=admin.pwd 
    --host=serverhost --config=rp --vs=rp --uri-prefix=// 

    To redirect to a secure site, follow the same step and provide the https address for the --server option.

    See CLI Reference, create-reverse-proxy(1).

  9. Create an instance for the rp configuration.

    wadm>create-instance --config=rp <host-name>

  10. Start the instances:

    wadm>start-instance --config=config1 <host-name>

    wadm>start-instance --config=rp <hostname>

    The web application deployed in config1 can be viewed through rp instance.

    http://<rp instance hostname>:8081/test

    See CLI Reference, list-reverse-proxy-uris(1), set-reverse-proxy-prop(1), get-reverse-proxy-prop(1), forward-reverse-proxy-header(1), block-reverse-proxy-header(1), and list-reverse-proxy-headers(1)

Modifying Reverse Proxy Parameters

ProcedureTo Modify Reverse Proxy Parameters

  1. Using the Administration Console, select Configurations > Virtual Servers > Content Handling > Reverse Proxy tab.

  2. Click the URI button.

    You can edit the following parameters:

    • URI — The reverse proxy URI

    • Server URL — Comma separated URLs of the remote server. If multiple values are given, the server will distribute load among the specified servers.

    • Sticky Cookie — Name of a cookie that when present in a response, will cause subsequent requests to stick to that origin server.

    • Sticky URI Parameter — Name of a URI parameter to inspect for route information. When the URI parameter is present in a request URI and its value contains a colon `:' followed by a route ID, the request will "stick" to the origin server identified by that route ID.

    • Route Header — Name of the HTTP request header used to communicate route IDs to origin servers.

    • Route Cookie — Name of the cookie generated by the server when it encounters a stickycookie in a response. The routecookie stores a route ID that enables the server to direct subsequent requests back to the same origin server.

    • Rewrite Headers — Comma separated list of HTTP request headers.

Configuring Timeout Parameter in Reverse Proxy

  1. Using the Administration Console, select Configurations > Virtual Servers > Content Handling > Reverse Proxy tab.

  2. Click the URI button.

    A new window appears.

  3. Click the HTTP Client Configuration link.

    You can edit the Idle Timeout parameter. The default value is 300.

Customizing Reverse Proxy

You can configure conditional request processing in reverse proxy by manually editing the virtual server specific obj.conf file or through CLI. After the configuration changes are done, it is recommended to deploy the configuration and start the instance so that the changes are implemented.

wadm>deploy-config config_name

wadm>start-instance --config config_name hostname

See CLI Reference, deploy-config(1), start-instance(1)

Note –

The appropriate obj.conf file used by your virtual server should be modified. It can be <vs>-obj.conf or the default obj.conf, depending on the configuration.

The following examples discuss some of the possible configurations in Web Server.