Java Network Programming FAQ
Java Network Programming FAQ.
Copyright 1998 - 2000 David Reilly. All Rights Reserved.
Last modification date : April 27, 2000
This document contains the answers to commonly asked network programming
questions posed by Java developers. Copies of this document mirrored at other
sites may be out of date, please ensure that you're looking at a current
version. The latest version of this document can be found at http://www.davidreilly.com/java/java_network_programming/
Table of contents
Overview
Legal
Comments
Books
Websites
Basic programming questions
1. Socket Questions
1.1 Should I use ServerSocket or DatagramSocket in my
applications?
1.2 How do I get the IP address of a machine from its hostname?
1.3 How do I perform a hostname lookup for an IP address?
1.4 How can I find out who is accessing my server?
1.5 How can I find out the current IP address for my machine?
1.6 Why can't my applet connect via sockets, or bind to a local
port?
1.7 What are socket options, and why should I use them?
1.8 When my client connects to my server, why does no data come
out?
1.9 What is the cause of a NoRouteToHost exception?
2. HTTP Questions
2.1 How do I display a particular web page from an applet?
2.2 How do I display more than one page from an applet?
2.3 How can I fetch files using HTTP?
2.4 How do I use a proxy server for HTTP requests?
2.5 What is a malformed URL, and why is it exceptional?
2.6 How do I URL encode the parameters of a CGI script?
2.7 Why is a security exception thrown when using java.net.URL
or
java.net.URLConnection from an
applet?
2.8 How do I prevent caching of HTTP requests?
Advanced programming questions
3. Advanced networking concepts
3.1 How do I handle timeouts in my networking applications?
3.2 How do I control the amount of time a socket will linger
before resetting?
3.3 What does the java.net.Socket.setTcpNoDelay method do, and
what is Nagle's algorithm?
3.4 How do I implement a (FTP/HTTP/Telnet/Finger/SMTP/POP/IMAP/..../)
client/server?
3.5 How do I implement PING in Java?
3.6 How can I send/receive email from Java?
4. Remote method invocation
4.1 What is remote method invocation?
4.2 When should I use remote method invocation?
4.3 Why can't I access RMI from C++?
4.4 Why won't my RMI implementation compile under JDK1.1?
4.5 Why won't my RMI implementation run under Java 2?
5. Common Object Request Broker Architecture
5.1 What is CORBA?
5.2 When should I use CORBA?
5.3 What support does Java have for CORBA?
5.4 How do I start the CORBA nameservice for Java 2?
6. Servlets
6.1 What are Java Servlets?
6.2 What do I need to develop servlets?
6.3 Where can I get more information on servlets?
6.4 How does servlet performance compare to applets?
6.5 How does servlet performance compare to CGI?
6.6 Should I use single-threaded, or multi-threaded, servlets?
6.7 How do I send cookies from a servlet?
6.8 How do I read browser cookies from a servlet?
6.9 How do I make cookies expire after a set time period?
6.10 Why aren't cookies stored by my servlets accessible to my CGI scripts or
ASP pages?
6.11 How can I void a cookie, and delete it from the browser?
Overview
This document contains the answers to commonly asked network programming
questions posed by Java developers. If you have a question not addressed by this
document, feel free to ask. My email is jnpfaq@davidreilly.com.
Legal
This document is Copyright 1998, 1999 David Reilly. All Rights are Reserved.
No unauthorized mirroring or archiving of this FAQ is permitted, other than for
personal use such as offline reading. Print distribution, or distribution on
CD-ROM, WWW sites (personal or commercial), or other electronic distribution is
strictly prohibited without prior permission. For redistribution rights, contact
jnpfaq@davidreilly.com.
This FAQ is no way connected with two books (published by Manning &
O'Reilly) by the same name. Java is a trademark of Sun Microsystems, and this
document is no no way connected with Sun Microsystems.
Comments
This document is by no means complete, though hopefully it is error free :-).
Anyone wishing to submit a question, or provide comments, can send email to jnpfaq@davidreilly.com
Books
Most good Java books contain at least one chapter of information on
networking topics. However, for the network programmer, this is invariably not
enough information! A trip to your local bookstore or library should turn up a
few books on the topic. If not, the following resources might help you.
Java Network Programming
188477749x.l.gif (60873 bytes)Author
: Hughes, et al
ISBN : 188477749X
This book is one of the best books to begin network
programming in Java with. Not to be confused with an O'Reilly book of the same
name, Java Network Programming covers a wide range of topics, from datagrams
to sockets, from message streams and encryption to remote method invocation
(RMI), CORBA, Java servlets, and a heavy dose of cryptography and security.
Packed full of code, this is the ultimate reference. A second edition of this
book, covering Java Platform 2, has just been released, and is highly
recommended.
[more]
Java Network Programming
Author : Elliotte Rusty Harold
ISBN : 1565922271
O'Reilly & Associates, in 1997 published a book by the name of "Java Network
Programming". This book received favorable reviews in online stores like Amazon for the first couple of years,
but is now woefully out of date, and not really a good choice. I've flicked through the title, and it doesn't cover newer
topics like CORBA, new RMI features, or the revised servlet system development kit. A second edition of this title, bringing
it up to speed, is in order.
[more]
The Java Tutorial
Author : Mary Campione and Kathy Walrath
ISBN : 0201310074
Published by Addison-Wesley, and available on-line from Sun, the Java
Tutorial provides a great introduction to the Java language, as well as
advanced material on both basic networking topics, and advanced issues, such
as RMI/CORBA.
http://java.sun.com/docs/books/tutorial/
Websites
The official source of any Java related information is Sun Microsystem's Java
site, located at http://java.sun.com/. Other
recommended sites include the following :-
Site : Java Coffee Break
URL : http://www.javacoffeebreak.com/
Features many Java networking articles about advanced networking topics, such
as RMI & CORBA, as well as an introduction to networking tutorial.
Site : Java Networking FAQ
URL : http://www.io.com/~maus/jnetfaq.html
The Java Networking FAQ is an older FAQ, and doesn't cover many of the newer
JDK1.1 networking features. However, it's still a good reference for networking
information.
Site : JavaWorld
URL : http://www.javaworld.com/
JavaWorld is an online magazine, published monthly. Covers a wide variety of
topics, some of which are networking related.
Site : Jguru
URL : http://www.jguru.com/faq/Networking
Offers a collection of networking FAQs for Java.
Basic programming questions
1. Socket questions
1.1 Should I use ServerSocket or DatagramSocket in my
applications?
DatagramSocket allows a server to accept UDP packets, whereas ServerSocket
allows an application to accept TCP connections. It depends on the protocol
you're trying to implement. If you're creating a new protocol, here's a few tips
DatagramSockets communciate using UDP packets. These packets don't
guarantee delivery - you'll need to handle missing packets in your
client/server
ServerSockets communicate using TCP connections. TCP guarantees delivery,
so all you need to do is have your applications read and write using a
socket's InputStream and OutputStream.
1.2 How do I get the IP address of a machine from its
hostname?
The InetAddress class is able to resolve IP addresses for you. Obtain an
instance of InetAddress for the machine, and call the getHostAddress() method,
which returns a string in the xxx.xxx.xxx.xxx address form.
InetAddress inet = InetAddress.getByName("www.davidreilly.com");
System.out.println ("IP : " + inet.getHostAddress());
1.3 How do I perform a hostname lookup for an IP address?
The InetAddress class contains a method that can return the domain name of an
IP address. You need to obtain an InetAddress class, and then call its
getHostName() method. This will return the hostname for that IP address.
Depending on the platform, a partial or a fully qualified hostname may be
returned.
InetAddress inet = InetAddress.getByName("209.204.220.121");
System.out.println ("Host: " + inet.getHostName());
1.4 How can I find out who is accessing my server?
If you're using a DatagramSocket, every packet that you receive will contain
the address and port from which it was sent.
DatagramPacket packet = null;
// Receive next packet
myDatagramSocket.receive ( packet );
// Print address + port
System.out.println ("Packet from : " +
packet.getAddress().getHostAddress() + ':' + packet.getPort());
If you're using a ServerSocket, then every socket connection you accept will
contain similar information. The Socket class has a getInetAddress() and getPort()
method which will allow you to find the same information.
Socket mySock = myServerSocket.accept();
// Print address + port
System.out.println ("Connection from : " +
mySock.getInetAddress().getHostAddress() + ':' + mySock.getPort());
1.5 How can I find out the current IP address for my
machine?
The InetAddress has a static method called getLocalHost() which will return
the current address of the local machine. You can then use the getHostAddress()
method to get the IP address.
InetAddress local = InetAddress.getLocalHost();
// Print address
System.out.println ("Local IP : " + local.getHostAddress());
1.6 Why can't my applet connect via sockets, or bind to a
local port?
Applets are subject to heavy security constraints when executing under the
control of a browser. Applets are unable to access the local file-system, to
bind to local ports, or to connect to a computer via sockets other than the
computer from which the applet is loaded. While it may seem to be an annoyance
for developers, there are many good reasons why such tight constraints are
placed on applets. Applets could bind to well known ports, and service network
clients without authorization or consent. Applets executing within firewalls
could obtain privileged information, and then send it across the network.
Applets could even be infected by viruses, such as the Java StrangeBrew strain.
Applets might become infected without an applet author's knowledge and then send
information back that might leave hosts vulnerable to attack.
Signed applets may be allowed greater freedom by browsers than unsigned
applets, which could be an option. In cases where an applet must be capable of
network communication, HTTP can be used as a communication mechanism. An applet
could communicate via java.net.URLConnection with a CGI script, or a Java
servlet. This has an added advantage - applets that use the URLConnection will
be able to communicate through a firewall.
1.7 What are socket options, and why should I use them?
Socket options give developers greater control over how sockets behave. Most
socket behavior is controlled by the operating system, not Java itself, but as
of JDK1.1, you can control several socket options, including SO_TIMEOUT,
SO_LINGER, TCP_NODELAY, SO_RCVBUF and SO_SNDBUF.
These are advanced options, and many programmers may want to ignore them.
That's OK, but be aware of their existence for the future. You might like to
specify a timeout for read operations, to control the amount of time a
connection will linger for before a reset is sent, whether Nagle's algorithm is
enabled/disabled, or the send and receive buffers for datagram sockets.
1.8 When my client connects to my server, why does no data
come out?
This is a common problem, made more difficult by the fact that the fault may
lie in either the client, or the server, or both. The first step is to try and
isolate the cause of the problem, by checking whether the server is responding
correctly.
If you're writing a TCP service, then you can telnet to the port the server
uses, and check to see if it is responding to data. If so, then the fault is
more than likely in the client, and if not, you've found your problem. A
debugger can be very helpful in tracking down the precise location of server
errors. You could try jdb, which comes with JDK, or use an IDE's debugger like
Visual J++ or Borland JBuilder.
If your fault looks like it is in the client, then it can often be caused by
buffered I/O. If you're using a buffered stream, or a writer (such as
PrintWriter), you may need to manually flush the data. Otherwise, it will be
queued up but not sent, causing both client and server to stall. The problem can
even be intermittent, as the buffer will flush sometimes (when it becomes full)
but not other times.
1.9 What is the cause of a NoRouteToHostException?
Usually this means that there isn't an active Internet connection through
which a socket connection may take place, or that there is a nasty little
firewall in the way. Firewalls are the bane of users and developers alike -
while useful for security, they make legitimate networking software harder to
support.java/java_network_programming/jnpfaq.txt
Your best option is to try using a SOCKS proxy, or to use a different
protocol, like HTTP. If you still have firewall problems, you can manually
specify a HTTP proxy server (see section 2.4)
This is a common problem, made more difficult by the fact that the fault may
lie in either the client, or the server, or both. The first step is to try and
isolate the cause of the problem, by checking whether the server is responding
correctly.
2. HTTP Questions
2.1 How do I display a particular web page from an applet?
An applet can instruct a web browser to load a particular page, using the
showDocument method of the java.applet.AppletContext class. If you want to
display a web page, you first have to obtain a reference to the current applet
context.
The following code snippet shows you how this can be done. The show page
method is capable of displaying any URL passed to it.
import java.net.*;
import java.awt.*;
import java.applet.*;
public class MyApplet extends Applet
{
// Your applet code goes here
// Show me a page
public void showPage ( String mypage )
{
URL myurl = null;
// Create a URL object
try
{
myurl = new URL ( mypage );
}
catch (MalformedURLException e)
{
// Invalid URL
}
// Show URL
if (myurl != null)
{
getAppletContext().showDocument (myurl);
}
}
}
2.2 How do I display more than one page from an applet?
The showDocument method of the AppletContext interface is overloaded -
meaning that it can accept more than one parameter. It can accept a second
parameter, which represents the name of the browser window that should display a
page.
For example,
myAppletContext.showDocument (myurl, "frame1")
will display the document in frame1. If there exists no window named frame1,
then a brand new window will be created.
2.3 How can I fetch files using HTTP?
The easiest way to fetch files using HTTP is to use the java.net.URL class.
The openStream() method will return an InputStream instance, from which the file
contents can be read. For added control, you can use the openConnection()
method, which will return a URLConnection object.
Here's a brief example that demonstrates the use of the
java.net.URL.openStream() method to return the contents of a URL specified as a
command line parameter.
import java.net.*;
import java.io.*;
public class URLDemo
{
public static void main(String args[]) throws Exception
{
try
{
// Check to see that a command parameter was entered
if (args.length != 1)
{
// Print message, pause, then exit
System.err.println ("Invalid command parameters");
System.in.read();
System.exit(0);
}
// Create an URL instance
URL url = new URL(args[0]);
// Get an input stream for reading
InputStream in = url.openStream();
// Create a buffered input stream for efficency
BufferedInputStream bufIn = new BufferedInputStream(in);
// Repeat until end of file
for (;;)
{
int data = bufIn.read();
// Check for EOF
if (data == -1)
break;
else
System.out.print ( (char) data);
}
}
catch (MalformedURLException mue)
{
System.err.println ("Invalid URL");
}
catch (IOException ioe)
{
System.err.println ("I/O Error - " + ioe);
}
}
}
2.4 How do I use a proxy server for HTTP requests?
When a Java applet under the control of a browser (such as Netscape or
Internet Explorer) fetches content via a URLConnection, it will automatically
and transparently use the proxy settings of the browser.
If you're writing an application, however, you'll have to manually specify
the proxy server settings. You can do this when running a Java application, or
you can write code that will specify proxy settings automatically for the user
(providing you allow the users to customize the settings to suit their proxy
servers).
To specify proxy settings when running an application, use the -D parameter :
jre -DproxySet=true -DproxyHost=myhost -DproxyPort=myport MyApp
Alternately, your application can maintain a configuration file, and specify
proxy settings before using a URLConnection :
// Modify system properties
Properties sysProperties = System.getProperties();
// Specify proxy settings
sysProperties.put("proxyHost", "myhost");
sysProperties.put("proxyPort", "myport");
sysProperties.put("proxySet", "true");
2.5 What is a malformed url, and why is it exceptional?
When you create an instance of the java.net.URL class, its constructor can
throw a MalformedURLException. This occurs when the URL is invalid. When it is
thrown, it isn't because the host machine is down, or the URL path points to a
missing file; a malformed URL exception is thrown when the URL cannot be
correctly parsed.
Common mistakes include :-
leaving out a protocol (eg "www.microsoft.com" instead of
"http://www.microsoft.com/")
specifying an invalid protocol (eg "www://netscape.com")
leaving out the ':' character (eg http//www.micrsoft.com/)
MalformedURLException will not be thrown if :-
the host name is invalid (eg "www.microsoft-rules-the-world.com")
the path is invalid (eg "http://www.microsoft.com/company_secrets.htm")
2.6 How do I URL encode the parameters of a CGI script?
This is an important question, as many Java applications and applets interact
with server side applications, servlets, and CGI scripts. Let's take a look at
how URL encoding works first though.
A URL can be used to invoke a server side application or script's GET method.
The first part of the URL will be the name of the server side script, followed
by a question mark '?' character. After that will come the name of each
parameter, and '=' sign to separate name from value, and a '&' character to
indicate the next parameter. Here's a fictitious example.
http://www.yourwebhost.com/yourcgi.cgi?name=your%20name&email=email@email.com
We can't include spaces or other high/low ASCII values, so the space
character has been substituted for %20 in this example. Java provides a
URLEncoder class to do this for us - we need only construct the URL and pass it
to the URLEncoder. Here's a quick code example to demonstrate.
String encodedURL = "http://www.yourwebhost.com/yourcgi?name=" +
// encode as value may have spaces or other characters
URLEncoder.encode("david reilly" );
System.out.println ("Encoded URL - " + encodedURL);
2.7 Why is a security exception thrown when using
java.net.URL or
java.net.URLConnection from an applet?
Web browsers impose security restrictions on applets, which prevent applets
from establishing network connections to servers other than that from which they
were loaded. Like socket connections, HTTP connections will cause security
exceptions to be thrown. If you absolutely, positively, have to access other
hosts (and replacing your applet with a Java servlet is impractical), consider
using a digitally signed applet.
2.8 How do I prevent caching of HTTP requests?
By default, caching will be enabled. You must use a URLConnection, rather
than the URL.openStream() method, and explicitly specify that you do not want to
cache the requests. This is achieved by calling the
URLConnection.setUseCaches(boolean) method with a value of false.
Advanced programming questions
3. Advanced networking concepts
3.1 How do I handle timeouts in my networking
applications?
If your application is written for JDK1.1 or higher, you can use socket
options to generate a timeout after a read operation blocks for a specified
length of time. This is by far the easiest method of handling timeouts. A call
to the java.net.Socket.setSoTimeout() method allows you to specify the maximum
amount of time a Socket I/O operation will block before throwing an
InterruptedIOException. This allows you to trap read timeouts, and handle them
correctly. For an article on the subject, see http://www.javacoffeebreak.com/articles/network_timeouts/.
If you're trying to handle connection timeouts, or if your application must support earlier versions of Java, then another
option is the use of threads. Multi-threaded applications can wait for timeouts,
and then perform some action (such as resetting a connection or notifying the
user). However, you as a programmer should avoid writing complex multi-threaded
code - particularly in your clients. There's usually an easy way to encapsulate
the complexity of multi-threading, and provide a simple non-blocking I/O
version. An article that demonstrates this technique for connect operations is
available at http://www.javaworld.com/jw-09-1999/jw-09-timeout.html
3.2 How do I control the amount of time a socket will
linger before resetting?
When a socket wishes to terminate a connection it can "linger",
allowing unsent data to be transmitted, or it can "reset" which means
that all unsent data will be lost. You can explicitly set a delay before a reset
is sent, giving more time for data to be read, or you can specify a delay of
zero, meaning a reset will be sent as the java.net.Socket.close() method is
invoked.
The socket option SO_LINGER controls whether a connection will be aborted,
and if so, the linger delay. Use the java.net.Socket.setSoLinger method, which
accepts as parameters a boolean and an int. The boolean flag will
activate/deactivate the SO_LINGER option, and the int will control the delay
time.
3.3 What does the java.net.Socket.setTcpNoDelay method
do, and what is Nagle's algorithm?
This method controls the socket option TCP_NODELAY, which allows applications
to enable or disable Nagle's algorithm. Nagle's algorithm (described in RFC
896), conserves bandwidth by minimizing the number of segments that are sent.
When applications wish to decrease network latency and increase performance,
they can disable Nagle's algorithm. Data will be sent earlier, at the cost of an
increase in bandwidth consumption.
3.4 How do I implement a (FTP/HTTP/Telnet/Finger/SMTP/POP/IMAP/..../)
client/server?
Your first step towards creating such systems will be to read the relevant
Request For Comments (RFCs) document. Not sure which one? There are specific
search engines, such as http://www.rfc-editor.org/,
that will allow you to search for the name of a protocol, and to then read
relevant documents. These RFCs describe the protocol you wish to implement.
3.5 How do I implement PING in Java?
Java includes support for UDP and TCP sockets. PING requires support for the
Internet Control Message Protocol (ICMP). Your only choice (at the moment), is
to use native code, or to use java.lang.Runtime to execute an external ping
application. You won't be able to develop a 100% Pure implementation.
NB - A native implementation that uses the Java Native Interface (JNI) is
available for PING, in both English and Spanish. See http://www.geocities.com/SiliconValley/Bit/5716/ping/
for more details.
3.6 How can I send/receive email from Java?
You can choose to implement Simple Mail Transfer Protocol (SMTP), to send
email, and either POP or IMAP to receive email. However, an easier alternative
is to use the JavaMail API, which provides a set of classes for mail and
messaging applications. Royalty-free implementations of the API are now
available from Sun for SMTP, POP and IMAP - and many other mail systems are
supported by third-parties. For more information, visit the official JavaMail
page, at http://java.sun.com/products/javamail/.
4. Remote method invocation
4.1 What is remote method invocation?
Remote method invocation (RMI), is a mechanism for invoking an object's
methods, even though the object is executing on a foreign Java Virtual Machine (JVM).
RMI is similar to remote procedure calls (RPCs), but has an added advantage -
method signatures can contain Java objects as well as primitive data types. Even
objects that a foreign JVM has never encountered before can be used, so new
tasks and methods can be passed across a network.
4.2 When should I use remote method invocation?
Here's a few rules of thumb :
When you wish to execute code on remote systems (distributed systems)
When you have a network that has machines capable of supporting JVMs on
all machines you would wish to connect to RMI
When you don't want your RMI applications to be used from machines that
don't support JVMs, or from applications written in C++/Ada/Cobol/Fortran/[insert
non-java lang here]
4.3 Why can't I access RMI from C++?
Remote method invocation allows method signatures to contain Java objects,
and C++ isn't capable of executing Java bytecode. If your RMI system only used
primitive data types, you might be able to write a software bridge between the
two - but this isn't direct remote method invocation. You'd be better off
investigating CORBA.
4.4 Why won't my RMI implementation compile under JDK1.1?
Under JDK1.02, RMI implementations extend java.rmi.server.UnicastRemoteServer.
This changed in JDK1.1 - you should now extend
java.rmi.server.UnicastRemoteObject.
4.5 Why won't my RMI implementation run under Java 2?
If you're running the client or server with Java 2, then you'll need to
specify a security policy file, to prevent SecurityExceptions being
thrown. This policy file will allow your application to bind to a local
port (if a service), and to connect to remote hosts (if a client).
The following changes should be made when running the client/server :
java -Djava.security.policy=java.policy yourserver
You'll also need to create a policy file (if one does not already exist).
Here's a sample policy file that will allow you to accept conections from ports
higher than 1024, but connect to all ports as a client.
grant {
permission java.net.SocketPermission "*:1024-65535",
"connect,accept,resolve";
permission java.net.SocketPermission "*:1-1023",
"connect,resolve";
};
5. Common Object Request Broker Architecture
5.1 What is CORBA?
CORBA stands for Common Object Request Broker Architecture. CORBA is a
mechanism that allows applications to invoke object methods that will execute on
remote systems. CORBA isn't limited to any single platform, or language. CORBA
systems are written in C++, Ada, Java, and other languages. This makes it more
interoperable than remote procedure calls and remote method invocation.
5.2 When should I use CORBA?
Here's a few rules of thumb :
When you wish to use services that are running on remote machines
(distributed systems)
When you have a heterogeneous networking environment, containing systems
that aren't capable of supporting JVMs because a port is not yet available
When you want your applications to be accessed by other systems written in
C, C++, Ada, Cobol, and other languages (including Java)
5.3 What support does Java have for CORBA?
Third party class libraries allow you to write CORBA application clients and
services in earlier versions of Java, but support for CORBA is also included
with the newly released Java 2 Platform. Clients and services can be written
(though an extra download of the idltojava tool is required). For more
information on Java IDL, see "Java
and CORBA - a smooth blend".
http://www.javacoffeebreak.com/articles/javaidl/javaidl.html
5.4 How do I start the CORBA nameservice for Java 2?
The nameservice can be started by running the tnameserv command, which should
be located in your Java 2 binaries directory. Remember however that this service
will run continuously, so you'll need to run it in a separate console window.
6. Java Servlets
6.1 What are servlets?
Servlets are server-side Java applications, as opposed to client-side applets
or standalone applications. While servlets are compatible with many different
types of servers, typically they are used in web servers, as a replacement for
CGI scripts or Active-Server Pages (ASP).
Java servlets offer many advantages over other forms of server-side
processing. Apart from the obvious (they are written in the Java programming
language, a big plus after all), servlet based applications are far easier to
write than CGI scripts. There's no need to write code for parsing HTTP request
parameters, as this code is provided by the javax.servlet.http package. You have
access to the entire Java API, with no networking restrictions (making servlets
far more attractive than applets). There are also performance increases over CGI
scripts, as servlets persist over time, and do not create a new process for
every connection.
6.2 What do I need to develop servlets?
To develop servlets, you need a basic familiarity with Java I/O streams,
HTML, and the HyperText-Transfer Protocol (HTTP). You'll also need to download
the Java Servlet Development Kit (JSDK) which is freely available from Sun
Microsystems. Finally, you'll need a web-server that supports servlets, or a
servlet engine which augments your server's capability.
6.3 Where can I get more information on servlets?
Gamelan's TechFocus recently published a two-part introductory tutorial on
servlets, which I can highly recommend. For more information see http://www.gamelan.com/journal/techworkshop/
I also highly recommend the book, Developing Java Servlets, by James Goodwill. The publisher is Sams,
and the ISBN is 0672316005.
For more information, see
http://www.davidreilly.com/goto.cgi?isbn=0672316005
6.4 How does servlet performance compare to applets?
Client-side Java has been dogged by performance problems due to slow loading
times, older JVMs without JIT compilation, and inefficient coding. Applets have
gained a reputation of poor performance, which often gives Java itself a bad
name. Server-side Java, however, doesn't suffer from the same performance
problems. The speed of execution is much better, as the server administrator can
install more recent JVMs (some of which are optimized for server-side
processing). The load time is virtually instantaneous, as there is no network
latency to contend with. You're also in a better position regarding security
restrictions - servlets can establish network connections without the sandbox
problems that plague unsigned applets.
6.5 How does servlet performance compare to CGI?
Servlet performance is often far superior to CGI. Rather than forking a new
process for each request, a pool of threads can be used to execute servlet
requests. Servlets persist across connections, so there's no need to perform
initialization tasks repeatedly. This, when coupled with a fast JVM and servlet
engine, can offer better performance than CGI.
6.6 Should I use single-threaded, or multi-threaded servlets?
By default, servlets are multi-threaded. If you specifically require a
single-threaded servlet, you should implement the javax.servlet.SingleThreadModel
interface. This guarantees that no two threads will be operating on the same
instance of the servlet, but still allows concurrent execution of the servlet.
The servlet engine will maintain a pool of available threads, and create new
instances of the servlet as required. This can cause performance problems
however, and should be used sparingly except on low-traffic servers.
6.7 How do I send cookies from a servlet?
HTTP is a stateless protocol, which makes tracking user actions difficult.
One solution is to use a cookie, which is a small piece of data sent by a web
browser every time it requests a page from a particular site. Servlets, and CGI
scripts, can send cookies when a HTTP request is made - though as always, there
is no guarantee the browser will accept it.
Cookies are represented by the javax.servlet.http.Cookie class.
Cookie has a single constructor, which takes two strings (a key and a value).
// Create a new cookie
Cookie cookie = new Cookie ("counter", "1");
Adding a cookie to a browser is easy. Cookies are sent as part of a
HTTPServletResponse, using the addCookie( Cookie ) method. You can call this
method multiple times, but remember that most browsers impose a limit of ten
cookies, and 4096 bytes of data per hostname.
public void doGet (HttpServletRequest request, HttpServletResponse response)
throws IOException
{
response.addCookie(new Cookie("cookie_name", "cookie_value"));
}
6.8 How do I read browser cookies from a servlet?
Reading cookies from a servlet is quite easy. You can gain access to any
cookies sent by the browser from the javax.servlet.http.HttpServletRequest passed to the servlet's
doGet, doPost, etc methods. HttpServletResponse offers a method, Cookies[]
getCookies() which returns an array of Cookie objects. However, if no cookies
are available, this value may be null, so be sure to check before accessing any
array elements.
// Check for cookies
Cookie[] cookie_jar = request.getCookies();
// Check to see if any cookies exists
if (cookie_jar != null)
{
for (int i =0; i< cookies.length; i++)
{
Cookie aCookie = cookie_jar[i];
pout.println ("Name : " + aCookie.getName());
pout.println ("Value: " + aCookie.getValue());
}
}
6.9 How do I make cookies expire after a set time period?
Depending on how you use the data stored in a cookie, it is sometimes a good
idea to make the cookie expire. Since anyone using the browser will have the
cookie sent on their behalf, it may appear to be a legitimate user when in
actual fact it is not. This often happens in places like Internet cafes, school
or university computing labs, or libraries. If your cookie sends a user
identifier that facilitates access to sensitive data, or allows changes to be
made (for example, a web-based email service), then you should expire cookies
after a small time period. If the user keeps using your servlet, you always have
the option of resending the cookie with a longer duration.
To specify an expiration time, you can use the setMaxTime(int) method of
javax.servlet.http.Cookie. It takes as a parameter the number of
seconds before the cookie will expire. For example, for a five minute
expiration, we would do the following :-
// Create a new cookie for userID from a fictitious
// method called getUserID
Cookie cookie = new Cookie ("userID", getUserID());
// Expire the cookie in five minutes (5 * 60)
cookie.setMaxTime( 300 );
When the cookie is sent back to the browser, using
HttpServletResponse.addCookie(Cookie), it will only be returned by the browser
until the expiration date occurs. If you'd prefer, you can also specify a
negative value for setMaxTime(int), and the cookie will expire as soon as the
browser exits. Note however that not everyone will shutdown their browser, and
it might be available for minutes, hours even days. Finally, specifying a value
of zero will expire the cookie instantly.
6.10 Why aren't cookies stored by my servlets accessible to my CGI scripts or
ASP pages?
By default, cookies are accessible to every HTTP request for the current
directory, and any subdirectories. Now on most web servers and servlet engines,
servlets are located in a special directory. For example, when using
servletrunner (which ships with the Java Servlet Development Kit), servlets must
be invoked under the /servlet/ directory.
http://webserver/servlet/servletname
If you want your cookies to be accessible elsewhere, you must specify the
root path of your webserver, using the javax.servlet.Cookie.setPath( String )
method.
// Create a cookie for everyone to share
Cookie myCookie = new Cookie ("counter", "1");
// Set path for cookie
myCookie.setPath( "/" );
Once you've set the path, any script (for example, stored in /cgi-bin/)
can access the cookies stored by your servlets.
6.11 How can I void a cookie, and delete it from the
browser?
You can specify an expiration date, using the setMaxTime(int) method of
javax.servlet.http.Cookie. Specifying a expiration time of zero
will void the cookie, and delete it from the browser.
// Expire the cookie immediately
cookie.setMaxTime( 0 );
// Send cookie back to the browser to void it
response.addCookie(cookie);