Servlet Engine Installation Guide
findinsite-js is a Java servlet. It is a program that must be run in a server that supports Java servlets.
This usually means installing a servlet engine and connecting it up to your
Please refer to our servers page for details of servers
and servlet engines.
Please consult our Getting Started Guide for an overview
of how to get findinsite-js, set up your server, install and configure findinsite-js.
Get the findinsite-js software from the Download page.
After an important Technical Preamble, this page gives some advice on
installing a Java servlet engine and
connecting it up to your web server.
Some types of servlet engine can be used for free
for testing purposes on your own computer without necessarily connecting it up to a
full web server.
Once your servlet engine is installed, please see the
findinsite-js Installation Guide
covering typical servlet engines.
If you have any questions on servlet engine installation,
please contact sales/support.
Please read this section explaining the jargon carefully.
A servlet engine runs servlets like findinsite-js.
Servlets are programs that respond dynamically to user requests.
The findinsite-js servlet responds to search questions with a web page containing the results sheet of hits.
It is common to test findinsite-js on your own computer first before installing
it on your main web server.
When testing on your own computer, you do not need to use a full web address, such
Instead, on almost all computers a web address
of http://localhost/ refers to your own computer.
If clicking on http://localhost/ does not work,
then you do not have a web server installed (or enabled) on your own computer.
Don't worry - you can still test findinsite-js using a servlet engine web server,
but you will not be able to connect it up to your own test web server
(although you can install a web server for most types of computer pretty easily).
Servlet Web server
In addition to your main web server, each servlet engine also usually
provides its own web server so that it can respond directly to requests,
so you might have two (or more) web servers on your computer.
The Apache JServ servlet engine must be connected to the main Apache server because JServ
does not include a servlet web server.
Having two web servers on a computer is achieved using ports. Each web server
responds at a specific port. A port is just a number between 0 and 65535.
A standard web server always responds at port 80, so http://www.findinsite.info:80/
is exactly the same as http://www.findinsite.info/.
The web server provided by a servlet engine will use a different port number which you can usually
choose, eg a web server at port 8080 on your computer should respond to this URL: http://localhost:8080/. (If you are not at the
standard web port of 80, please choose a value in the range 1024 to 65535.)
findinsite-js on a Servlet web server
As an example, on a servlet web server, findinsite-js might be found at this URL:
In this case, the servlet web server responds at port 8111 on your computer
The last part of the URL,
/findinsite/servlet/findinsitejs, tells the servlet engine which servlet it should run.
During initial testing, you will probably run findinsite-js direct from the
servlet engine web server.
However, you will soon want to connect your main web server to
your servlet engine. When connected up, your main server will divert some requests to
be handled by the servlet engine, eg all requests that start
might be passed to the servlet engine.
Using this example, this request to your test main server
asks the servlet engine to run findinsite-js.
Your main web server and the servlet engine are usually connected together using a port.
This port number is usually different from the servlet web server port number,
so the connector port number might be 57000.
When connected to the main web server, you may well decide to disable the servlet web server, so that
the servlet engine only responds to requests that come from the main server via the connector.
Example final set up
In a typical findinsite-js scenario, two computers are set up:
your own computer for testing, and your main web server.
In some cases, eg on an Intranet, these may be the same computer.
While you can usually set up your own computer, only your site's webmaster or your ISP
may be able to set up your main server.
Here is an example of a typical final set up.
Most links are likely not to work now.
See below for possible alternative URLs.
Macromedia JRun usually supplies two servlet web servers,
so the Test and/or Main server computers will usually have another "Admin" web server
at a different port.
Servlet Engine Installation
Installing a servlet engine is usually fairly painless. It is a two stage process:
- Install the servlet engine. Test that you can run servlets.
- Connect the servlet engine to your main web server.
Test your servlets again.
Servlet Engine Installation
Obviously, each Servlet Engine is different, and therefore the installation instructions
A servlet engine needs a Java runtime environment. The servlet engine may come with a built-in
Java runtime environment, or you may need to get some more software.
A servlet engine that runs JSP may also need a Java compiler - again this may or may not be included.
You usually just install the servlet engine and its servlet web server first.
You can usually delay connecting up the servlet engine to your main web server until later.
Some servlet engines have nice installation programs that set everything up for you.
In other cases, you will have to edit one or two set up files before the software
will run. In particular, you may need to choose (or change) the port number for the
servlet web server and the connector. You may also need to set the server name.
In some cases, you may need to set environment variables on your computer.
Some servlet engines can be run in different ways. In Windows NT/2000/XP/Vista/7 some servlet engines
can be run as a service, which means that they start whenever the computer starts.
An alternative is to run the servlet engine as an ordinary program, which means you have start it
by hand, eg from the Windows Start menu. If you make changes to the servlet environment,
eg by adding findinsite-js or by altering initialisation parameters, then you may have to restart
the servlet engine. Being able to stop and start the servlet engine easily is useful when
testing. Some tools provide a web interface that makes it easy to
restart the servlet engine.
Once you have installed your servlet engine, check that you can access its web server home page,
eg if you installed the server at port 8111 on your own computer, this URL should
Have a look round the supplied web pages, and check that any supplied servlets run satisfactorily.
You can now try to connect the servlet engine up to your main server.
Or you can leave it until later
and jump straight to the findinsite-js Installation Guide.
If you have one of these Java servlet engines installed, then
the relevant page will describe more servlet engine configuration options.
Connecting your Servlet Engine to your main server
As described above, you need to connect your servlet engine
to your main web server for it to process findinsite-js search requests.
Again, each Servlet Engine and Server is different, so there are quite a few permutations
of installation instructions. Please consult the
servlet engine documentation for full details.
You may need to stop your main web server before attempting to connect to it.
Once connected up, you typically need to restart your main server.
Some servlet engines provide tools to automate the process of connecting to your web server.
In Macromedia JRun, the "JRun Management Console" automates most of the tasks; however for at least
one server type, the Microsoft IIS server, we found that we needed to consult the
documentation for the final connection steps.
For other servlet engines, such Apache JServ, the main Apache server configuration files need to be altered
considerably. To make it easier, JServ provides example files that can be pasted into the
Apache server configuration files, with little or no alteration.
In almost all cases you will need to specify a connector port number.
See above for an explanation of ports and a connector.
Similarly, you will usually need to specify a connection point for the servlet engine.
A connection point is sometimes called a context path. You could make the
/servlets so that all requests to the main server that have
at the start of the path are sent to the servlet engine for processing.
As an example, if you specified
/servlets as the connector context for the findinsite-js
servlet engine on your test computer, then http://localhost/servlets/
will display the servlet engine home page. With the same context path,
will usually run findinsite-js.
In some cases, the connector context is empty, so findinsite-js would appear at
Please see the note below for possible alternatives to these URLs.
There are many advanced ways to configure a servlet engine,
and its connection to the web server.
For a start, you may want to turn off the servlet engine web server functionality
and only allow access via the main server connector.
Other possible options include increasing the memory allocation pool,
setting up multiple servlet contexts,
multiple private or shared Java Virtual Machines or supporting virtual hosts.
You can usually configure multi-threading support and session timeouts.
Please see your servlet engine documentation for details of these tasks.
Your choice of server, servlet engine and context path might change the URLs that you have to
use to run findinsite-js. You might also need to adjust
some settings in the findinsite-js configuration screen.
In all the examples below, a host name of
localhost has been used.
If you are setting up your main server, then substitute your web domain name instead, eg
Contact sales/support if you need further help running findinsite-js.
The findinsite-js welcome page has a search form for several common types of servlet engine.
See these pages for likely URLs and any post-installation changes required
for these types of servlet engine:
As described above, your computer may have a main web server and a servlet web server
running at the same time, each using different ports. In a URL, the path after the host name may be different
for each web server.
For example, suppose a servlet engine has a web server at port 8111 on your own computer.
If the context path for the findinsite-js Web Application is
/findinsite then you should
be able to access findinsite-js using this URL:
(If you used a context path of
/haggis for the findinsite-js Web Application, then
findinsite-js would appear at http://localhost:8111/haggis/servlet/findinsitejs.)
Now, suppose the servlet engine is connected to your main server via context path
then findinsite-js will be visible on the main server at this URL:
For some server connectors, the connector context path is empty,
so findinsite-js would appear on the main server at: