Pro Programmers

java.lang.ClassFormatError: Truncated class file=javax.servlet.ServletException

2011-10-18T17:09:00.001-07:00

Problem

Having above exception after deploying WAR file.

Solution

Delete [Tomcat]\work\Catalina\localhost.

Cause

Tomcat’s got crazy with JSP compilation.

Spring Bean as Servlet Filter

2011-08-17T19:42:00.001-07:00

Wish

Want to use a Spring bean as JSP Filter.

Solution

Use DelagatingFilterProxy from Spring libarry.

Details

Implements the Spring bean from javax.servlet.Filter as usual Filter

public class HeaderValidationFilter implements Filter{

……

}

Create Spring bean (with Spring XML) as usual

<bean id="filterHeader” class=”com.proprogrammers.HeaderValidationFilter”>

……

</bean>

In “web.xml” add JSP filter as below:

Notice that <filter-name> is the same as bean id above.

<filter>
    <filter-name>filterHeader</filter-name>
    <filter-class>org.springframework.web.filter.DelegatingFilterProxy</filter-class>
</filter>

<filter-mapping>
    <filter-name>filterHeader</filter-name>
    <url-pattern>*</url-pattern>
</filter-mapping>

<servlet>
    <servlet-name>Spring</servlet-name>
    <servlet-class>org.springframework.web.servlet.DispatcherServlet</servlet-class>
    <load-on-startup>1</load-on-startup>
</servlet>

Caveat

Filter bean (at step-2) cannot be defined in Spring Servlet XML file; Spring-servlet.xml in the sample. It must be in other Spring XML file.

http://static.springsource.org/spring-security/site/docs/3.0.x/reference/security-filter-chain.html

Credits

Collin Yates

http://forum.springsource.org/showthread.php?20230-Howto-The-joy-that-is-DelegatingFilterProxy

[Java] How to Make Asynchronous Method

2011-08-01T19:04:00.001-07:00

I have a long-running method which I want to run asynchronously. I transform the method into asynchronous method by running it on separate thread.

Original Code:

    public void mainMethod() throws InterruptedException{
        System.out.println("Main Method start");
        longRunningMethod("testing", new Object());
        otherMethod();
    }

    public Integer longRunningMethod(String param1,Object parm2) throws InterruptedException{
        System.out.println("Long running method");
        Thread.sleep(5000);
        //long process
        return 1;
    }

    public void otherMethod(){
        System.out.println("Other Method");
    }

New Code:

public void mainMethod() throws InterruptedException{
        System.out.println("Main Method start");
        //longRunningMethod("testing", new Object());
        asyncServiceMethod("testing",new Object());
        otherMethod();
    }

    public Integer longRunningMethod(String param1,Object parm2) throws InterruptedException{
        System.out.println("Long running method");
        Thread.sleep(5000);
        //long process
        return 1;
    }

    public void otherMethod(){
        System.out.println("Other Method");
    }


    public void asyncServiceMethod(final String parm1,final Object obj){
        Runnable task = new Runnable() {

            @Override
            public void run() {
                try {
                   longRunningMethod(parm1,obj);
                } catch (Exception ex) {
                    //handle error which cannot be thrown back
                }
            }
        };
        new Thread(task, "ServiceThread").start();
    }

There are few things to consider wrapping a method into asynchronous call.

Passing Parameters must be final
Exception from the long-running process cannot be thrown back to mainMethod (main thread). It has to be handled at separate thread (ServiceThread in the sample)
Returning value cannot be retrieved (Integer value in this sample) from the main Method. (If one wants to retrieve, FutureTask should be used.)

MS SQL – Strip Time, Day, Month from DateTime

2011-05-09T18:24:00.001-07:00

In the database, Date Time values is stored to millisecond level and for many times, I need to group record by Day, Week or Month or Year and I want to get day only or start day of the week or start day of the month, etc. Here I just put a note on SQL for the purpose mentioned.

select getdate() as CurrentDateTime;

select DATEADD(d,DATEDIFF(d,0,getdate()),0) as CurrentDate;

select DATEADD(wk,DATEDIFF(wk,0,getdate()),0) as CurrentStartDayOftheWeek;

select DATEADD(mm,DATEDIFF(mm,0,getdate()),0) as CurrentStartDayOftheMonth;

select DATEADD(yy,DATEDIFF(yy,0,getdate()),0) as CurrentStartDayOftheYear;

Result

CurrentDateTime
-----------------------
2011-05-10 11:18:54.857
CurrentDate
-----------------------
2011-05-10 00:00:00.000
CurrentStartDayOftheWeek
------------------------
2011-05-09 00:00:00.000
CurrentStartDayOftheMonth
-------------------------
2011-05-01 00:00:00.000
CurrentStartDayOftheYear
------------------------
2011-01-01 00:00:00.000

Java byte Array into int Array

2011-05-03T18:14:00.001-07:00

public int[] toIntArray(byte[] barr) {
        //Pad the size to multiple of 4
        int size = (barr.length / 4) + ((barr.length % 4 == 0) ? 0 : 1);

        ByteBuffer bb = ByteBuffer.allocate(size *4);
        bb.put(barr);

        //Java uses Big Endian. Network program uses Little Endian.
        bb.order(ByteOrder.LITTLE_ENDIAN);


        int[] result = new int[size];
        bb.rewind();
        while (bb.remaining() > 0) {
            result[bb.position()/4] =bb.getInt();
        }

        return result;
}

Disappointed for not seeing shifting? Wanna complain about performance?

This is definitely not the most efficient method to convert byte array into int array yet it can be done with simple logic and a bit of knowledge in NIO.

Unless the program is to be used on 1K memory device, using ByteBuffer is acceptable I reckon.

(I tried to convert to IntBuffer but in vain. If anyone could point out, that’d be great.)

Peter has mentioned how to use with IntBuffer in comment section as follow:

public int[] toIntArray(byte[] barr) {
        //Pad the size to multiple of 4
        int size = (barr.length / 4) + ((barr.length % 4 == 0) ? 0 : 1);

        ByteBuffer bb = ByteBuffer.allocate(size *4);
        bb.put(barr);

        //Java uses Big Endian. Network program uses Little Endian.
        bb.order(ByteOrder.LITTLE_ENDIAN);
        bb.rewind();
        IntBuffer ib = bb.asIntBuffer();
        int [] result = new int [size];
        ib.get(result);

        return result;
}

BigDecimal equals method

2011-03-01T05:49:00.001-08:00

BigDecimal class from java.lang.Math package is useful for scale manipulation. Be careful though when you want to compare two BigDecimal objects with equals() method since it tests not just the equality of value but also the scale as mentioned in the document as below:

…..this method considers two BigDecimals equal only if they are equal in value and scale (thus 2.0 is not equal to 2.00 when compared by this method).

For the following code

BigDecimal a = new BigDecimal("2.0");
BigDecimal b = new BigDecimal("2.00");
if (a.equals(b)){
System.out.println("Equals");
}else{
System.out.println("Not Equals");
}

it will return "Not Equals” as the scales of the objects are different.

To compare values of the above two objects, compareTo method can be used as it implements Comparable interface like wrappers.

if (a.compareTo(b)==0)
{
System.out.println("They have same value");
}

Chrome Proxy Setting

2010-11-23T20:17:00.001-08:00

If you change proxy setting in Chrome using Options->Under the Hood –> [Change Proxy Setting], it will use IE proxy setting and it will change proxy setting for all other browsers.

To avoid this, execute chrome.exe using the command line argument below:

--proxy-server=PROXYSERVER:PORT

[Spring] log4J

2010-11-04T17:44:00.001-07:00

WEB-INF\web.xml

<context-param>
        <param-name>log4jConfigLocation</param-name>
        <param-value>/WEB-INF/log4j.xml</param-value>
    </context-param>
    <listener>
        <listener-class>org.springframework.web.util.Log4jConfigListener</listener-class>
    </listener>….

WEB-INF\log4j.xml

<?xml version="1.0" encoding="UTF-8" ?>
<!DOCTYPE log4j:configuration SYSTEM "log4j.dtd">

<log4j:configuration xmlns:log4j="http://jakarta.apache.org/log4j/">
    
    <appender name="FILE" class="org.apache.log4j.DailyRollingFileAppender">
        <param name="file" value="../logs/MyApp.log" />
        <param name="datePattern" value="'.'yyyy-MM" />
        <param name="append" value="true" />
        <layout class="org.apache.log4j.PatternLayout">
            <param name="ConversionPattern" value="%d [%t] %-5p %c{6} - %m%n"/>
        </layout>
    </appender>

    <root>
        <priority value ="info" />
        <appender-ref ref="FILE" />
    </root>

</log4j:configuration>

Enable Ping (Windows 2008)

2010-10-31T22:01:00.001-07:00

According to WiKiPedia

Ping is a computer network administration utility used to test the reachability of a host on an Internet Protocol (IP) network and to measure the round-trip time for messages sent from the originating host to a destination computer.

The name comes from active sonar terminology.

Ping is used to test the hostname/IP address for DNS resolution.

Ping uses ICMP protocol while telnet uses TCP protocol on specific port.

Since Windows 2008, it is disabled by default. To enable it, Networking – Echo Request(ICMPv4In), from command prompt, enter following:

netsh firewall set icmpsetting 8

[MSSQL] Reseed Identity

2010-10-25T17:04:00.001-07:00

dbcc checkident( 'TABLENAME', RESEED, 200000000)

Details in: http://blogs.techrepublic.com.com/datacenter/?p=406 (Susan Harkins, 2008)

[Spring] Initialization Callbacks

2010-09-20T18:28:00.001-07:00

If you want the Spring container to perform initialization work on your bean after all the properties are set, you may implement InitializationBean and implements afterPropertiesSet() method.

But just as Spring documentation outlines, that approach will couple the code to Spring. You should use init-metod instead and call any public method from your bean. You could event pass parameters using constructor-arg.

http://static.springsource.org/spring/docs/2.0.x/reference/beans.html#beans-factory-lifecycle

Tomcat 5.5 with Java 1.6

2010-09-02T17:55:00.001-07:00

If your Tomcat is running with Java 1.5 by default and want to upgrade to Java 1.6, double click TOMCAT\bin\tomcat5w.exe

Change Java Virtual Machine by selecting ..\Java\jdk1.6.0_xx\jre\bin\server\jvm.dll.
There is one more trick to make it work. Copy ..\Java\jdk1.6.0_xx\bin\msvcr71.dll into the same folder as jvm.dll.

Thanks to kellyohair http://forums.java.net/jive/message.jspa?messageID=155421 for step-2

Enable connection pooling in Spring Datasource

2010-07-14T18:38:00.001-07:00

I use Apache Commons DBCP for the above purpose

<bean id="poolDataSource" class="org.apache.commons.dbcp.BasicDataSource" destroy-method="close">
     <property name="driverClassName" value="net.sourceforge.jtds.jdbc.Driver"/>
     <property name="url" value="${url}"/>
     <property name="username" value="${username}"/>
     <property name="password" value="${password}"/>
     <property name="poolPreparedStatements" value="true"/>
     <property name="maxActive" value="10"/>
     <property name="maxIdle" value="10"/>
</bean>

Caused by: org.xml.sax.SAXException: unable to find FieldDescriptor for 'elementName' in ClassDescriptor of ParentElement

2010-07-09T07:46:00.000-07:00

I got an error similar to the title in the course of XML binding Castor. It happens when there is namespace declared in XML document just like

<country xmlns="http://pro-programmers.com/schema">
   <city>London</city>
</country>

For the above XML document mapping was designed as

<mapping xmlns="http://castor.exolab.org">
   <class name="com.proprogrammers.pojo.Country">
   <map-to xml="country" />
   <field name="city" type="string">
   <bind-xml name="city" node="element"/>
   </field>
   </class>
</mapping>

Solution: I declared namespace in binding xml element as below:

<map-to xml="country" />

<bind-xml name="city" xmlns="http://pro-programmers.com/schema" node="element"/>

</field>

</class>

</mapping>

See more here.

Log File

2010-06-22T22:44:00.001-07:00

Application Log files should

roll over daily or by size

Manual/Scheduled process should be involved to archive log files after a period (for instance every month). It is necessary to set log file archiving procedure.

analysable

for number hits/requests (even if the application is not web app, it should log how many times a module is called). We keep such log in a separate log file – say “access.log”.
performance profiling - We use perf4j and log the performance in “performance.log” which can be enabled as needed.
Database requests - Not every developer can tell how many DB access application submit in a certain time or for a request. This is very helpful in optimising performance.

follow industrial standard

such as Apache log format so that it can be analysed with tools available (e.g; Web Log Expert)
Even if the log file is not for web application, it may be formatted in Apache log format to be analysed with tool.
Log file may be formatted in CSV sometimes.

be Visible for Errors

We keep any application error in a separate log file, “error.log” which improves visibility and we have a monitor program to raise alarm if any given exception is thrown.

include start-up states

log file should have application restart time,
which modules are running and which are not
which configurations are used
Operation/Production team who starts the system should be able to view the log clearly and should be fully aware of how application is started and how it will behave after start-up.

be configurable

from INFO to DEBUG without restarting the system

be Compact

There were times when application server froze since there was no space left in the HDD. Developers log whatever the want and as a result log file is bloated with unnecessary statements. Apart from consuming storage space, bloated log file reduce the visibility.
Code review should involve checking log statement.

Conceal sensitive data

Data like password or credit card number should never be in the log file.

viewed by Log file monitoring tool

We use BareTail to monitor log file which is very good especially with large size file and searching log statement in regex

(This post will be updated as I can think of any standard for log file in our team…)

Credit Card Track2 Data

2010-06-22T18:15:00.001-07:00

Remove Transfer-Encoding:chunked from HTTP Response Header

2010-05-06T22:53:00.001-07:00

Note From Wikipedia

Chunked Transfer Encoding is a mechanism that allows HTTP messages to be split in several parts. This can be applied to both HTTP requests (from client to server) and HTTP responses (from server to client).

HTTP Chunked Transfer Encoding allows a server to maintain a HTTP persistent connection for dynamically generated content.

Problem

If HTTP Response has “Transfer-Encoding” header, it causes client to wait for closing the connection.

Cause

If there is no “Content-Length” is defined in HTTP response header, it will be chunked transmission. HTTP Response must have either “Transfer-Encoding” or “Content-Length” attribute.

Solution

Define “Content-Length” attribute if it can be done.

HibernateSystemException: could not set a field value by reflection setter

2010-04-08T18:27:00.001-07:00

The above error is caught when NULL value in a database column is attempted to set to a setter in an Entity bean whose attribute mapped to that column type is of primitive type.

Never ever use primitive for attribute type in Entity Bean.

Since Java 5, with a feature called “Autoboxing”, we can seamlessly treat wrapper as primitive type as below

Integer deptId;
deptId = 5;
//deptId = new Integer(5);

int a;
a = deptId;
//a = deptId.intValue();

Compiler will handle it.

Preventing NullPointerException

2010-04-06T19:28:00.001-07:00

We have several ways to avoid getting NullPointerException caught such as a method returning empty array or List rather than NULL as following:

public String[] getNames(){
return new String[0];//return null
}

This code will have the method caller in peace who can just write code without having to worry about NULL value something like this:

for(String name:getNames()){
//Do Something
}

But there could be a scenario where we want to have empty value different from NULL in return, for instance:

public String getName(){
return null;
}

For that reason, we can never assume that a method could not return NULL. Whether a method returns NULL or empty value should be described in Javadoc I believe. Consider this code:

if(person.getName().equals("John")){
//Do Something
}

The above code will give NullPointerException when getName() returns NULL and it should be written as below if we know that NULL can be returned.

if ("John".equals(person.getName())){
//Do Something
}

The conditional clause above will yield false if getName() returns NULL.

Zip Vs GZip

2010-02-28T16:53:00.001-08:00

GZip	Zip
compresses single file (can concatenate the contents to compress multiple files) ideal for only one data stream	compresses multiple files ideal for archiving files
deflates the content	tries to choose which one will product the smallest output (Unix uses “deflate” algorithm)
Java classes: GZipOutputStream, GZipInputStream	Java classes: GZipOutputStream, GZipInputStream

What is faster: System.currentTimeMillis() or new Date().getTime()?

2009-11-23T20:41:00.001-08:00

For micro-optimisation or trivial memory utilization improvement, System.currentTimeMillis() is faster and better. The constructor in the Date class uses System.currentTimeMillis().

public Date() {
this(System.currentTimeMillis());
}

Thus, just to get a long value for current time in millisecond, it’s good to use System.currentTimeMillis().

long start = System.currentTimeMillis();
//process
long elapsed = System.currentTimeMillis() - start;

Above code saves garbage collector from cleaning up a couple of unnecessary Date objects.

If you want more precision, use System.nanoTime() which returns the current value of the most precise available system timer, in nanoseconds. This is especially useful when the elapsed time is resulted as zero millisecond in the above code.

Database Diagram support object cannot be installed

2009-11-19T15:56:00.001-08:00

Problem: When using Microsoft SQL Server Management Studio to create database diagram, I got this error message:

Solution: I run the following query:

EXEC sp_dbcmptlevel 'DB_NAME', '90';
go
ALTER AUTHORIZATION ON DATABASE::DB_NAME TO "LOGIN"
go
use [DB_NAME]
go
EXECUTE AS USER = N'dbo' REVERT
go

where DB_NAME and LOGIN are variable. In my case, LOGIN is NT User name since I’m using Windows Login.

Creating ThreadPool in Spring

2009-09-22T23:50:00.001-07:00

<bean id="threadPoolExecutor" class="java.util.concurrent.ThreadPoolExecutor" init-method="prestartCoreThread">
    <constructor-arg index="0" type="int" value="16"/>
    <constructor-arg index="1" type="int" value="256"/> 
    <constructor-arg index="2" type="long" value="1"/>
    <constructor-arg index="3" type="java.util.concurrent.TimeUnit" >
        <bean class="java.util.concurrent.TimeUnit" factory-method="valueOf">
            <constructor-arg>
                <value>HOURS</value>
            </constructor-arg>
        </bean>
    </constructor-arg>
    <constructor-arg index="4" type="java.util.concurrent.BlockingQueue">
        <bean class="java.util.concurrent.SynchronousQueue"/>
    </constructor-arg>
</bean>

[JSP] Include : Directive Vs Action

2009-07-19T18:51:00.001-07:00

[JMeter] – Max Threads with Rump up and Loop Count

2009-07-02T19:11:00.001-07:00

Synopsis

Wish to simulate a test plan with, say, 3000 or 4000 threads per minute.

So JMeter is set up as above.

Problem

Every time it is run, JMeter stops at around 2500 threads. It looks like JMeter has max thread count.

Causes

JMeter does not have thread limit but JMeter creates all the thread (4000 in this case) up-front before it starts running the first thread. When it attempts to create 4000 thread, System (OS) might not allow to create that high number of thread. JMeter creates as much as it is allowed and that’s the reason JMeter stops at around 2500 threads.

In fact JMeter run all the threads it has created; it just couldn’t create the number of threads specified. Number of max thread JMeter can created can be checked with JConsole monitor at the execution time.

Practice

Use Loop count instead rather than using high number of threads.

Solution

To achieve the same requirement it is configured like following:

That will allow us to send 400 requests in 6 seconds and will repeat for 10 times and 4000 request will be sent after a minute.

Catch

If we simulate high number of hits with Loop count, requests are not distributed evenly across the Ramp-up time.

Following result is from the test plan configured as :
Number of Threads = 6
Ramp-Up in sec = 60
Loop Count = 5

Following result is from the test plan configured as :
Number of Threads = 30
Ramp-Up in sec = 60
Loop Count = 1

This illustration can help us clearly see how 30 requests are distributed in a minute.

Final Solutions

If we want to simulate high number of test cases (samples) then it is better we use Constant throughput Time or some other type of Timer which allow us to simulate, for instance, 6000 hits per minute for the whole day by creating only one thread. How? Find out yourself.

Pro Programmers

java.lang.ClassFormatError: Truncated class file=javax.servlet.ServletException

Problem

Solution

Cause

Spring Bean as Servlet Filter

Wish

Solution

Details

Caveat

Read More

Credits

[Java] How to Make Asynchronous Method

MS SQL – Strip Time, Day, Month from DateTime

Java byte Array into int Array

BigDecimal equals method

Chrome Proxy Setting

[Spring] log4J

WEB-INF\web.xml

WEB-INF\log4j.xml

Enable Ping (Windows 2008)

[MSSQL] Reseed Identity

[Spring] Initialization Callbacks

Tomcat 5.5 with Java 1.6

Enable connection pooling in Spring Datasource

Caused by: org.xml.sax.SAXException: unable to find FieldDescriptor for 'elementName' in ClassDescriptor of ParentElement

Log File

Credit Card Track2 Data

Remove Transfer-Encoding:chunked from HTTP Response Header

HibernateSystemException: could not set a field value by reflection setter

Preventing NullPointerException

Zip Vs GZip

What is faster: System.currentTimeMillis() or new Date().getTime()?

Database Diagram support object cannot be installed

Creating ThreadPool in Spring

[JSP] Include : Directive Vs Action

[JMeter] – Max Threads with Rump up and Loop Count

Synopsis

Problem

Causes

Practice

Solution

Catch

Final Solutions