Tutorials & Help

There would be times when we forget things which we have been using for a long time. Then you would want to go to basic tutorial and read all about it. There is no good re-writing the commands and explanations here in blog but it is good to have links to the tutorials and help documents. This post will contain links to well-written help documentation for everything I will need. Please contribute to it, if you wish.

1. JAR basics - Documentation from Oracle
   
2. Tutorial on Marathon testing tool
3. How to install Eclipse plugins - manually and the easy way?
   
4. How to install Kindle for Linux
   

Ubuntu: Terminal display help

Here are the few tricks up your sleeve for utilizing the scripts to customize the terminal display.

1. To have the shell prompt always start at the left hand side of terminal:

When you have long pathname for the current working directory and you have a even longer command to execute, the display gets all screwed up. So it always helps you to have the shell prompt start at left hand side of terminal. You have to change the PS1 - to customize your bash prompt.

In the fig, you can't distinguish where the dir name ends and where the command starts. So open ".bashrc" file in home folder

gedit ~/.bashrc

and add the following code. Note: as a good coding practice, add a line of comment saying when and why added the piece of code and add the new code at the end of the file.
PS1=$'\[\e]2; \h::\]$PWD\[\a\]\[\e]1;\]$(basename "$(dirname "$PWD")")/\W\[\a\] \w\n\T \$ '

Then the bash prompt becomes like this,
The current working directory name in first line(s) and in a new line current time followed by $ for you to type in the command.

Much better huh, what exactly does the code do? Let's review from back
\$ - if the effective UID is 0, a #, otherwise a $
\T - the current time in 12-hour HH:MM:SS format
\n - newline
\w - current working directory with ~ instead of $HOME directory
\[ \] - begin and end of a series of non-printing characters
\W - basename of the current working directory with ~ instead of $HOME directory
\e - escape character
\h - hostname

Source: to customize you bash prompt you can follow this help document.

You can also create aliases for opening and refreshing bashrc file like this (if you have to change it frequently)

In "~/.bashrc" file add the following code,

alias refprofile=' source ~/.bashrc'
alias editprofile='vi ~/.bashrc'

whereas, editprofile - opens the file for editing and refprofile - refreshes the file to reflect the change made.

Happy customization!

Ubuntu: Set up iSCSI initiator and target

There will be a lot of posts about iSCSI hereafter, as I am working on the SAN protocol for my graduate project. This was the source I used : http://www.howtoforge.com/using-iscsi-on-ubuntu-9.04-initiator-and-target

This is what I did:
Follow the set up till creating logical volume.
If your system got a partition (/dev/sdan - where n is any number between 1 and 9) which is free and doesn’t contain a partition table(no data or os in it). Use Gparted to format the disk to any type. Then follow the steps below:

sudo fdisk -l(lower case L) - This will list the device partitions size and ID

pvcreate /dev/sda3 - I’m using this partition - On success it would say “Physical volume /dev/sda3 successfully created

vgcreate vg0 /dev/sda3 - On success “Volume group vg0 successfully created”. vg0 - name of the volume group. Any format is fine.< lvcreate -L20G -n storage_lun1 vg0 - From the source 3. After creating the logical volume, follow the source instructions. At the place of iqn.2001-04.com.example:storage.lun1 192.168.0.100 substitute your IP address. I’m going to have both target and initiator at the same computer for now. So I’m using 127.0.0.1 in place of 192.168.0.100 in all forthcoming commands. However, 3260 port number is same and it is default for iSCSI.

------------------------------------------------------------------------------------------------------------------------------------
Now the above post is good but once I partition the sdb disk and changed some file I had problem logging to target. This is the error I got.

iscsiadm: Could not login to [iface: default, target: iqn.2001-04.com.example:storage.disk2.amiens.sys1.xyz, portal: 192.168.1.101,3260]: iscsiadm: initiator reported error (15 - already exists) iscsiadm: Could not log into all portals. Err 15.

I tried couple of solutions like deleting the devices modules and taking a look at the /var/log/messages. Then I jumped at this page and all I had to install tgtadm service and it was piece of cake from there and simple solution to create small portions to act as targets. Since our project is about able to replay the iscsi event with session details as original event for problem diagnosis, for testing during development we need to create small partitions. I would keep updating about this project. If you need more details, do drop me an email.

Ubuntu: How to create full system backup

Since I'm playing around my computer to install iSCSI target and initiator, I needed to create volume groups. That involves partitioning and formatting of my hard drive. Hence I'm very sure this would result in the loss of files from my hard drive :) So I wanted to take full system back up of my Ubuntu. I found a very useful and simple to-do list here. I am going to list those steps here FMR(For My Reference).

The general idea in taking full system back up in Ubuntu(or any Linux for that matter), you can create one compressed archive of all the files with folder structure,then uncompress the archive when you want and viola you have saved your files! Unlike Windows, the creating back up and restoring can be done when the system is running. Warning: while restoring we have to be careful where we uncompress the files to avoid losing data. Here are the steps:

• Become a root user using "sudo su". Warning: be cautious when you are a super user and execute a command from online, there is a single command which can destroy the entire filesytem!
• Go to root of the filesystem using "cd /". This would be the place where your back up archive file would be created. Feel free to create the file in remote drives directly.
• Then use the following single command to create the backup archive
  

  tar cvpzf backup.tgz --exclude=/proc --exclude=/lost+found --exclude=/backup.tgz --exclude=/mnt --exclude=/sys /

  
  This file would be really big and if you want to use better compression technique, you can also create bzip2 files.
  

  tar cvpjf backup.tar.bz2 --exclude=/proc --exclude=/lost+found --exclude=/backup.tar.bz2 --exclude=/mnt --exclude=/sys /

  
  This commands take a long time to create the file. The explanation for the command and the options:
  • c- compress, v- verbose, p- preserve permissions
    
  • backup.tar - filename for the backup file
  • You need to exclude the directories that doesn't contain useful files and the back up file itself. Remove any /mnt if you need to avoid backing up other mounted partitions too.
    
  • Ignore the error message like "tar: /: file changed as we read it
    tar: Exiting with failure status due to previous errors" which you might get at the end of the process.
  
  
• Restoring:
  The command to restore the file from the same place as root of the filesystem,
  

   tar xvpfz backup.tgz -C /

  
  Make sure you create the folders that you excluded from the back up archive. So this would be the list of directories if you had used the above command:

  mkdir proc
  mkdir lost+found
  mkdir mnt
  mkdir sys
  etc...

  
  Then reboot the system to complete the process.

Thanks to the source: http://ubuntuforums.org/showthread.php?t=35087

Ubuntu: Configuring Webcam and Microphone in Lenovo Y530

I have a Lenovo Y530 laptop. I had Windows then being the devoted Linux girl, I installed Ubuntu 9.10 on it. The laptop has built-in webcam(1.3M pixels). It was working fine with windows but got screwed up in ubuntu. So I set out to find out how to fix it. At first glance it looked like there was no proper solution available online. But after hours of search, I found the easy steps buried under a bunch of useless/complex solutions. So I wanted to document the exact steps I followed to configure the webcam.

• I tried installing cheese(webcam booth tool for linux) and used Ekiga softphone. You can use either of this tool to check if the webcam is listed in video devices list. In Cheese, the screen had this test screen and both the tools din't recognize the webcam.
• Then in Terminal, try "lsusb". It would list all the USB devices connected to the computer. A sample lsusb would look like this (this has no webcam)
  

  
  Bus 005 Device 002: ID 0483:2016 SGS Thomson Microelectronics Fingerprint Reader
  Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
  Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
  Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
  Bus 002 Device 003: ID 046d:c016 Logitech, Inc. M-UV69a/HP M-UV96 Optical Wheel Mouse
  Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
  Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
  

  The 8 digit number followed by ID is the unique ID of the hardware device. If at all you want to search for solutions for any device use this number in your searches and you will more likely land in the page you want. When I ran this cmd, the output din't have any device listed in it. Then in some thread in Ubuntu forums(an excellent place to start your search), someone asked to use hot keys to toggle the webcam. In Lenovo Y530, its Fn+Esc. After that, lsusb listed the device as in,

  Bus 002 Device 002: ID 04f2:b105 Chicony Electronics Co., Ltd

• Now, I searched the ID with all related keywords, it landed here. I checked for my device and saw that it was complaint and all i have to do is download the v4l dvb drivers and viola.
• Then follow the steps listed in this link. I downloaded the latest version of V4L/DVB via web browser and following cmds "make", "make install". [don't forget to do, "sudo make install" - I know its simple but trust me you don't want to miss it]
  

Then happened the weirdest thing. After rebooting the system and checking if webcam worked with cheese, the microphone got screwed. When open the sound preferences dialog(System->Preferences->Sound), in the Input tab, there was no device listed. Without proper knowledge about the thread flow, I removed pulseaudio from the system. Then it was a fresh start as with the webcam. The steps followed were the following:

• In Terminal, use this cmd

  lspci -v | grep -i audio

  
  The result would print out the Audio device if configured right. In my case, it returned empty result.
• So I downloaded the alsa drivers, util, lib files from here. Followed this steps
  • copy the tar.gz files to "/usr/src/alsa" - if you have the folder already or create one if you don't
  • Then unzip the files using "sudo tar xjf alsa-driver*". Similarly util and lib files(change tar options as you require)
  • Then we have to go each directory and do "./configure", "make", "make install".
  • During the "./configure" if any error pops, you have to clear that for the installation to complete. I am just listing the problems I faced.
    • Problem: During the "make" of Util files
      

       xmlto man alsactl_init.xml
      /bin/bash: xmlto: command not found
      

      
      Solution:

      sudo apt-get install xmlto

    • Problem: During the "./configure" of Util files

      checking for new_panel in -lpanelw... no
      configure: error: panelw library not found

      
      Solution: Create soft links to the library files. click here for source

       sudo ln -s libpanelw.so.5 /usr/lib/libpanelw.so
      sudo ln -s libformw.so.5 /usr/lib/libformw.so
      sudo ln -s libmenuw.so.5 /usr/lib/libmenuw.so
      sudo ln -s libncursesw.so.5 /lib/libncursesw.so

    
    

Then reboot the system. Now everything works perfectly fine.

Later added: As I mentioned in the post, I accidentally removed pulseaudio from my system. So when I tried to do skype call, both sides webcam worked fine. But audio had too much static noise. So I had to install pulseaudio. Then reinstalled skype. Skype call works fine. But still if I use the system's music player, then try skype calls again audio has static noise. I read from a forum that "sudo killall pulseaudio" and then restarting skype helps. But that is not efficient solution. If someone has permanent solution to this please drop a comment. Thanks in advance.

Servlets

Servlets are replacing CGI script at most enterprise computing. They both perform similar operations of creating and displaying HTML pages dynamically. The basic differences between the two:

• CGI - Platform dependent, hence they cannot utilize server's capabilities; spawns new process for every request, hence resource intensive;
• Servlets - Since written in Java they can be executed in servers without any modifications as per "write once, run anywhere"; spawns new threads within the server process, hence lite weight.

Lifecycle of servlets:

• init() - initializes the servlet
• service(ServletRequest , ServletResponse) - all url's access this method and here we delegate to other service methods of form doXXX(HttpServletRequest , HttpServletResponse)
• destroy() - destroys the servlet

Reset root password for phpmyadmin/mysql

I recently faced an issue where I forgot the root password for mysql and ended up searching for simple procedure to unlock it by resetting the password. So I want to save here the steps to unlock the phpmyadmin if I should loose my root password again. Then I can skip the searching-for-an-hour part and go straight to the unlocking part. (Thanks to fredanthony and for source thread: click here)

The error message you get when you enter wrong password in the login page is :

"#1045 - Access denied for user 'root'@'localhost' (using password: NO) "

Steps:

1. Stop mysql - $ /etc/init.d/mysql stop
2. Kill all mysql related processes. The ps list can be pretty big and confusing so better use grep to filter mysql processes and use only those pids in kill command. - $ ps waux | grep mysql, $kill -9 pids
3. Open a new session of mysql by skipping permissions - $ /usr/bin/mysqld_safe --skip-grant-tables
4. While the above session is open, in a new tab open a new session of mysql - $ /usr/bin/mysql
5. Inside the new session,change the database - (inside mysql)$use mysql;
6. The critical part of our unlocking process is resetting the password for root user -(inside mysql) $UPDATE user SET Password=PASSWORD('YOUR_PASSWORD_HERE')
   WHERE Host='localhost' AND User='root'; - This step sets new password for the root user.
7. Then quit the mysql sessions and restart the mysql - $quit, $/etc/init.d/mysql restart
8. try logging into phpmyadmin with new password you set in update command and viola!
   

P.S: List of useful SQL command syntax - source

METHOD 2:
This step is to change the MYSQL root password. Source: Ubuntu help documentation. In a terminal enter the following command:

"sudo dpkg-reconfigure mysql-server-5.1"

However verify you have mysql-server-5.1 in your system. To find that, in terminal type "mysql -V". The result would print out details about the mysql server installed in your system.Sample result was
"mysql Ver 14.14 Distrib 5.1.41, for debian-linux-gnu (i486) using readline 6.1". Hence I use 5.1 in the above command. When you run the reconfigure command, it will prompt you to enter password and restarts the mysql daemon.

Ubuntu: Tips and tricks

An excellent and powerful feature about Ubuntu or any other Linux machines is that you can do anything/everything with a single command in a terminal. As with great powers comes great responsibility of understanding(possible) and remembering(absolutely impossible) those powerful commands! So this post would have simple and sometimes complex(which I doubt) commands for working your way around Ubuntu. Please contribute to this list, if you wish. Thanks in advance.

commands:
/tr>

Install eclipse	sudo apt-get install eclipse
Uninstall eclipse	sudo apt-get autoremove eclipse
Copy entire directory in Ubuntu	cp -r src_path/folder/* dest_path
Install LAMP stack	sudo apt-get install lamp-server^ - ^ cadet symbol is mandatory
Restart/Start/Stop Apache	sudo /etc/init.d/apache2 restart/start/stop
Install phpmyadmin	sudo apt-get install libapache2-mod-auth-mysql phpmyadmin
Cleaning up after any uninstall	sudo apt-get autoremove
Start/Stop MySql	/etc/init.d/mysql start/stop
Create softlink for phpmyadmin in Ubuntu 10.04(access as http://localhost/pma/index.php)	sudo ln -s /usr/share/phpmyadmin /var/www/pma
Open folder as root user - AVOID this and prefer using terminal and sudo for executing commands	gksudo nautilus
Print out processor information	cat/proc/cpuinfo
To find out if 32 bit or 64 bit linux kernel	uname -a - if result has i386/i486/i586/i686 GNU/LINUX, then 32 bit or x86_64 GNU/Linux, then 64 bit
To view .chm files in Ubuntu - chm = Microsoft compiled HTML files format	Just to read use "xchm"(sudo apt-get install xchm). To edit the file and republish in different format, follow the source here.
To find the version of JDK in your linux	"java -version"
To create aliases for often used complex commands	Cmd: alias alias_name="alias_cmd --options".If you got more aliases and you need it for all sessions hereafter, create a file "./alias" (name is for your reference, any name should do) in home directory. Add all the aliases command you need.Add the following lines in file $HOME/.bashrc, # Source the aliases, if a separate file exists if [ -e $HOME/.alias ]; then [ -n "$PS1" ] && . $HOME/.alias fi and execute "source $HOME/.bashrc". Source here.
To change password for a user account	cmd: "passwd". You will be prompted to enter current, new and retype new password. It should be more than 6 char long.
To change login shell	Find the current shell, "echo $SHELL", the env variable has the current shell info. Then find list of valid shell logins for your system by "cat /etc/shells". Finally to change to new shell, "chsh -s newshell" where newshell is the full pathname taken from /etc/shells. You will be prompted for password. It will take effect from next login.
To clear the contents of a file without deleting it	We know "cat filename" - print the file on terminal, "nano filename" - to edit line by line, "> filename" - clear the contents of the file without destroying the file.
To use gcc-3.x version instead of default gcc in Ubuntu in 10.04	You need to download required 3.x version, I need 3.4 so I downloaded gcc-3.4-base, cpp-3.4, and gcc-3.4 Then "sudo dpkg -i *.deb", this would select the removed databases and install the gcc. Your default gcc would still be the same and both version co-exist. While compiling just use "gcc-3.4". Note: "gcc -v" displays the version of gcc.
To execute user defined script from any directory in any session	Lets say you got a script named "first" and you want to execute the script just like other linux commands like "echo" from anywhere at shell prompt. In short, you need to add your bin directory to the PATH env variable. Thats what exactly we do here: 1. "mkdir bin" --> in your home directory create a directory called bin . Its a good practice to use "bin" folder to store your scripts. 2. "cp first ~/bin" --> copy all your tested script files to the bin directory. 3. "cd" --> go to root directory, thats where you can find .bash_profile. 4. "vi .bash_profile" --> open the profile in vi editor. This file has all startup settings. 5. "export PATH=$PATH:~/bin" -->append the path of bin directory to env variable PATH so system can find the executable. Don't forget the ":" and replace "~" with full path to your directory if you are using other directory than your home. Save the vi file using ":wq". 4. "first" --> it should work from any directory of the user.
How to install apache server	In terminal: "sudo apt-get install apache2". This would install,configure the server for you. To check in browser, "http://localhost"
How to open "rar"/"zip" files	1. First install "sudo apt-get install unrar". Then "unrar x filename.rar" where filename.rar = your file2. To unzip files, in terminal "unzip filename". -d option specifies the target directory for the files.
Create alias ip address in Linux	"ifconfig INTERFACE:ctr IP_ADDR netmask 255.255.255.0 up"Exp: INTERFACE - the interface in which you want to add the new ip address. ctr - start from 0 and increment for every new alias ip. IP_ADDR - the new alias ip address you want to add. netmask, up - keywords mandatory. Calculate the netmask and update it. If i want to add 192.168.30.10 to eth1 the command would be "ifconfig eth1:0 192.168.30.10 netmask 255.255.255.0 up".
To display unix timestamp	"date" displays the date and time. If you want the unix timestamp use "date +%s". Eg: For Mon Jan 24 21:42:57 PST 2011, the timestamp is 1295934173. Use this site to find out timestamp given date and time.
How to find version of Tomcat	1. "cd $TOMCAT_HOME/bin" -->Go to TOMCAT_HOME - root folder of tomcat installation 2. "sh version.sh" -->will print out the version details of tomcat, jdk
How to open new terminal and execute commands from a terminal	The command is "screen". In the old terminal, "screen command_to_execute" will open up terminal inside this one. Look here for options to detach the new terminal.

P.S: Usually blogger has a bugging problem of inserting unwanted spaces if you insert a table in Edit Html section. To avoid the pitfall, after pasting the table code, remove all the spaces. Hence the code like this (avoiding printing it as table)

table
tbody
tr
td /td
tr
/tbody
table

After removing the extra spaces would look like this,

table tbody tr td /td /tr /tbody /table

Now the code looks confusing, so better edit the table with spaces and finally before publishing it remove the spaces. Hope this helps someone!

To-Do: Run a marathon

Just like every other Sunday morning, I started the hike in Mission peak. The trial never gets easy even if you hike them every week. The climate was awesome as usual. While we were climbing down the trail we saw the runners of 23rd Annual Ohlone 50K Wilderness Run

Its a 31 mile run for some good cause. I saw people from all age groups run for the cause. When I saw them, I added a new item to my bucket list[list of things I plan to do before life ends] and that is to run a marathon for a cause. And I really do hope I am able to do it not just once but almost every year. I will make it happen soon!! For now I will just add my small achievement of touching the mission peak here.

View from Mission peak

Looking back at the trail

PHP MySql POST GET Cookies Session - Bits and information

I always feel its easier to remember details if we learn it as differences between two objects. So I want to blog the basic differences between cookies and session, POST and GET method. And bits and pieces of information about UNION,UNION ALL and some more. This post has so many random information. Let me know if you find something wrong or different.

Three ways to move information across web pages:

1. links: a link with href set to destination page.
   
2. Forms: set action to point to target, method of how server should process information sent, encoding type optional. Both 1 and 2 requires user action like a click.
   
3. Without any user action, user header in PHP to move to another page. header("Location: page.php");
   

Cookies:

• Cookies are stored at the client's computer.
• Once stored at client comp by an application, it can be accessed by any web page of that application.
• It contains variable = value pairs of information.
• Cookies are more useful for applications that has no database.
• Cookies will be available from the next page and not in the current page where it was set.
• In PHP, setcookie('variable','value'); sets the cookie and if you don't mention time, this cookie will expire when user exits the application.
• On setting the cookie, it gets stored in one of many built in arrays of PHP called $_COOKIE. For accessing your cookie variable just refer the array with the variable name like $_COOKIE['variable']
• If we want our cookie to be present even after user exits application then we have to mention the expiration time while we set the cookie. We can use two functions along with setcookie() - time and mktime. time() - returns the current time, we have to add the seconds for which the cookie needs to be alive. eg. setcookie('var','val',time()+3600) - to set cookie alive for 1 hour(60*60 seconds). mktime() - returns date and time and the order of arguments passed is hr,min,sec,mon,day, year. eg. setcookie('var','val',mktime(3,0,0,5,6,2010)); - to make it expire by 3am of May 6th 2010.
  
• To remove cookies, use the same setcookie without any values passed. eg. setcookie('var'); setcookie('var','');
  
• setcookie limitation: It has to be before any output is sent to the client browser. This limitation is with both session and cookie.
  

Session:

• Session details are stored in a file at the server side. In Unix and Linux systems, in \tmp folder and in Windows folder called sessiondata. We can change the location of where the file is stored by chaning session.save_path value in php.ini.
  
• Session ID - long nonsense number for every client which cannot be guessed or forged. In PHP, system variable is used for session ID - PHPSESSID.
• Session ID is passed to every page to access the session details. There are three ways to pass them. 1. If cookies turned on, use the cookies. 2. For links, use the URL. 3. Use hidden variables for form with POST method.
• Session variable is got and stored in $_SESSION built in array(similar to $_COOKIE array in cookies). They are accessed same way as in cookies.
• For session to work we need to enable, track_vars while installing PHP.(default its turned on from PHP 4.0).
• If cookie is turned off at the client side, trans_sid should be enabled to transfer session id. To enable that use session.use_trans_sid = 1 in php.ini.
  
• Start a session: session_start() - If sessionID is found, then load $_SESSION with variables and their values. If no ID is found, then it is first time so create a new session and set PHPSESSID.
• Save a session variable: $_SESSION['var'] = 'val';
• Close a session: session_destroy() - destroys the session details.
• session_id() - returns the PHPSESSID value - current session id.
• unset($_SESSION) - unset the session details in the current page.
• Limitation: session has to be set before any output is sent to the client browser.
  

We know that POST and GET are ways of indicating how the server should process the information sent by the form.

POST:

• Sends information in 2 steps. 1. browser contacts form processing server specified in action. 2. Once contact has been established, send data to server in separate transmission.
• On server side: 1. read parameters from a standard location. 2. After read, decode parameters before application use form variables.
• To get post variables from earlier form use built in array $_POST.
  

GET:

• In a single transmission, data is sent to the server. Data is appended by the browser to action URL.
• This is the default method if not specified otherwise.
  
• On server side, gets information passed at the end of the URL.

POST or GET what to use?

• GET: best transmission performance(single transmission) and apt for small forms with short/few fields.
• POST: Apt for forms with many/long text fields.
• If inexperienced with server programming use GET to avoid extra steps of processing- read/decode as in POST.
• For security purpose, use POST to avoid the information you transmit to be available in open for hackers to track. POST has security holes too but atleast it has encoding when transmitting.
• To invoke server processing outside form tag, eg in a tag, use GET because it lets us use form-like parameters as part of URL.
  

ENCTYPE field in FORM tag:

• Two types of enctype options available. 1. Multipart and 2. Text/plain
• Multipart: forms with file selection fields for upload by user.
• text/plain: used along with mailto in action attr of form tag. While sending forms to email server rather than a server.
• The default encoding type: Internet media type Application/x-www-form-urlencode.

UNION and UNION ALL:

• Union all - combines rows from multiple row sources into one result set. It includes duplicates.
• Union - does the same thing as union all but excludes duplicates, result would be sorted here without duplicates.

Polymorphism in OO langauges:

• Ability of 2 or more objects belonging to different classes to respond to exactly same message in different class specific ways.
  

Thread programming in Java

I spent a day in library refreshing stuffs I learned long back(not so long!) while I was in undergrad. Since my memory is fresh with these topics, I thought I should blog them so next time I can find it all in one place. On a different note, a day spent studying/reading in library is well spent! Okay back to Threads in Java.

Thread - class, Runnable - interface which has the thread implementations. For our application class to use thread we need to either implement or subclass or make the class a member class of our application. Thus three ways to create threads:

1. Extend Thread class: For standalone applications that don't need to extend another class.
2. Implement Runnable interface: For classes that need to extend other classes. Due to single inheritance, we can achieve threads by implementing the interface. Eg: Applets that extend JApplet and can't extend Thread too.
   
3. Construct a thread passing an inner class that is Runnable: For threads with tiny run methods and little outside interaction.

Thread life cycle functions:

• Thread() - When subclassing Thread class
• Thread(name) - when subclassing + for debugging
• start() - Begins process of assigning CPU time to a thread. Finally it results in calling run() method of the thread object.
• suspend(), resume(),stop() - deprecated. Because these suspend and resume functions can cause deadlocks.
• interrupt() - interrupt whatever the thread is doing
• boolean isAlive() - returns true if thread is neither finished or terminated by call to its stop()
• join() - to wait for a thread to finish
• get/setPriority() - higher priority threads gets first chance of CPU
• wait(), notify(), notifyAll() - semaphore and for thread synchronization.
• synchronised - keyword - can be used for any object in Java. When used over an object or a method, threads will wait if one thread has already started running.

Thread synchronization:
The wait, notify and notifyAll methods allow any java.lang.Object to be used as a synchronization target.

wait():
This causes the current thread to block in given object until awakened by notify or notifyAll.

notify():
Randomly selected thread waiting on this object is awakened. Then that thread tries to regain the monitor lock. If wrong thread is awakened, will result in deadlock.

notifyAll():
All threads waiting on this object is awakened. Then all try for the monitor lock. Hopefully one of them succeed.

Best practices:

• To avoid deadlock, use notifyAll to wake up waiting threads.
• To avoid using deprecated method - stop, use the variable as a flag and use it to find if the thread is done executing.
  

Today's scribbling

I took a tamil book from Sunnyvale Library for my Mom but I ended up drawing the art that was on the cover! And of course I am reading the book too. Here is my reproduction of the art :)

While I was at it, I drew something for my niece Ovia too. Her favorite character of the bunch, Goofy.

There are lot more art on the book's cover, I will post them later!! Ciao.

Who said I can't touch Mission Peak?

Well, I have this habit of proving myself to be what I claim, when somebody challenges me. So thats what happened when my uncle said I can't hike all the way to the Mission peak and I did it. Its this mountain hiking place in Milpitas. My uncle hikes the place with his pals every Saturday. I am an outdoorsy person. But my uncle had hard time believing it and challenged me to touch the peak this sunday(April 11th, 2010). Hence started one of my memorable experience.

I used to go trekking every summer when I visit here while I was in undergrad. But that was just simple not-so-steep mountain climb. This one was longer and steeper than the usual place in Los Altos. We reached the place on time and started climbing. It was steep right from the foot of the mountain. To be honest initially it was tough, but not something I can't handle. We saw a "big" rabbit sleeping on the grass near the trail, light drizzle and so many beautiful scenes as we hiked. The peak of this experience is the peak itself. We couldn't have picked a day with such worse weather. Though it was so mild and fun at the bottom as we reached peak, wind was very strong. It literally pushed us through in some places, I am not exaggerating. The peak was so beautiful and was worth the pain and cold. The fun part was when we started climbing down. It was more difficult than climbing up because it was very difficult to put a break. If you loose a minute of your body control and let your legs loose you will snowball down to the bottom of the mountain, well may be not in good shape because the trail bends in place!!

It took 3 hours for the climb but as I said, it was all worth it. Now, my family agrees I am an outdoorsy person. I am planning to continue this awesome hiking every weekend in this summer!! But I seriously hope the wind becomes weaker and it is not so cold in the coming weeks.

P.S: I touched the Mission Peak proof is the picture above.

POSIX Thread programming in C

I have to work on a project for Network programming subject this semester. The project objective is to develop an application that acts as a TCP server on top of UDP socket. Hence our project should take care of everything a TCP server basically and internally does for us, right from three way handshake till 4 way handshake. While there is that, the server have to be concurrent and it should handle few attacks like xmas tree, syn flooding.

So for the concurrency in server handling, our team decided to go with thread instead of forking new process to handle incoming connections since thread is lighter than forking. But the real reason being is I have worked on process creation in Unix for some projects in Undergrad. But I never experimented with thread creation. So what is learning when there is no experimentation involved.

While I am at it, I thought I should also blog the things I learn from thread programming in C. These things usually tend to slip away with time. Hence the post. Now this post may contain very basic instructions but for me these are baby steps. So if you are looking for complex details, this is not the post you should be reading. Thanks.

Misc:

• Good sources: source1, source2, source3
• All thread programs need "#include " and while compiling use "gcc -pthread filename.c" to include pthread.h. Hence use IDE to write code but terminal to run the files.

Virtual OS

I never really appreciated the use of virtual OS softwares until I had to carry two laptops around all day just because the simulator I need for the network lab assignment can run only on windows. So here I am installing VMware in Ubuntu 9.10 followed by Windows XP and over it the famous OPNET. I am charting the steps for my future reference and to help if someone happen to stumble upon my blog. I am also chart out other miscellaneous things that I had to learn when I do this.

Miscellaneous:

1. How to find RAM on ubuntu ?

gnome-system-monitor &

This system monitor provides better visualization of the system resources and its usage ( Source)


2. How to run .rpm file in Ubuntu?

• sudo apt-get install alien
• alien -k name_of_rpm_file.rpm
• dpkg -i file.deb

The "alien" converts the rpm file into its .deb equivalent. Then run the .deb file to install the package which we downloaded as .rpm file. (source)

3. How to remove non empty directory in linux?

rm -rf

This forces and does remove non empty directory recursively. Caution while removing root directories. (source)

4. How to run .deb file in linux?

dpkg -i name.deb

It installs the software from the package. (source)

5. What is difference between .rpm and .deb installations? which one is preferred?

RPM and DEB are both package types. While using RPM, we have to take care of all dependency between files in the installation process. In DEB it is taken care and we have to just use the above command and install the package. RPM can be very frustrating and time consuming. (source)


Main VMware installation: (source1) (source2)

1. Download .tar.gz file from vmware site and also register to get the license details.
2. Uncompress the file using "tar -zvxf VMware-server-2.0.2-203138.i386.tar.gz"
3. go to the directory "vmware-server-distrib" and execute "./vmware-install.pl"
4. It looks like we need to download the patches to install vmware separately. We can get that using "wget http://www.ubuntugeek.com/images/vmware-server.2.0.1_x64-modules-2.6.30.4-fix.tgz"
5. Uncompress the file using "tar xvzf vmware-server.2.0.1_x64-modules-2.6.30.4-fix.tgz" and execute the shell as super user using "sh vmware-server.2.0.1_x64-modules-2.6.30.4-fix.sh"
6. To avoid directory confusions, we need to remove a directory as super user using "rm -rf /usr/lib/vmware/modules/binary"
7. After installing the patch required, we should be able to run "./vmware-config.pl" command successfully.
8.Then follow the instructions in the installation.
9. Continue with the default values specified for the network connections, NAT creations.
10. After a lot of probing and poking, it will ask for the activation license serial number. You would have it in a link in the registration mail. Then follow the on screen instructions.
11. If everything goes successful, you will get the following screen


Starting VMware services:
Virtual machine monitor done
Virtual machine communication interface done
Virtual ethernet done
Bridged networking on /dev/vmnet0 done
Host-only networking on /dev/vmnet1 (background) done
DHCP server on /dev/vmnet1 done
Host-only networking on /dev/vmnet8 (background) done
DHCP server on /dev/vmnet8 done
NAT service on /dev/vmnet8 done
VMware Server Authentication Daemon (background) done
Shared Memory Available done
Starting VMware management services:
VMware Server Host Agent (background) done
VMware Virtual Infrastructure Web Access
Starting VMware autostart virtual machines:
Virtual machines done

The configuration of VMware Server 2.0.2 build-203138 for Linux for this
running kernel completed successfully.


12. Now we have to manage the server using firefox browser. In address space type https://:8333 for HTTPS and http://:8222 for HTTP.

13. After adding the certificate security, we need to login as root. If we don't have a password set for root, we can do it using "sudo passwd root". For this to happen, user should have administrator privileges.

14. Login using the newly set root password.

15. Use the interface to install new OS and continue working on OS over OS..

16. Create a virtual machine with options as Windows XP, NAT, 256MB RAM, 8GB memory, with CD and USB linked to host operating system.

17. Insert Windows(in my case) installation cd in the disc drive and when you power on the virtual machine, it would detect automatically and start the installation process. I noticed that this process was far more easier than installing windows on a real machine.

18. Install VMware tools to get good responsiveness.

19. Enjoy working on two OS seamlessly.

Happy Coding!!


LATER ADDED:

God it feels so good to have both the OS in one laptop and to swtich between them easily. Anyways I found one other excellent guide to do the same as I stated above. It just uses apt-get to install vmware server. But I prefer, when you want to learn, do it by downloading package and installing it. You will learn more than you signed up for. But in future when you are just reinstalling, you can just follow the simple steps as here. (source)

Art - Love of my life

Its been a long time, I posted something non technical and about my interests. Well, the title is pretty clear about what I am gonna talk about in this post. I always had a knack for painting and all the artsy stuffs. Given that nobody in my family's been interested in art, this is something that I can proudly say I immensely enjoy learning with an open mind. As one of my art teacher told me, art is all about having an open mind and different perspective on a scene. All this started when I very badly wanted to do something for school annual day competitions. I had a friend who was very good at drawing things as such, I guess I can say I may be inspired by him. When I won the drawing competition, I felt so happy and started doing more paintings. Till now I have done quite a number of drawings and few paintings. I love

1. drawing cartoons:
2. Glass painting:
3. Oil painting:
4. Coffee painting (Monochromatic painting):
5. And ofcourse trying out pencil shading and other stuffs.

I did a lot of glass paintings in college which I gifted my friends and professors. Even last week when I talked after one year to my research professor for whom I gave a picture as our farewell gift, he said everybody loved it. These little things make me happy and content with my ability.
Recently since I am busy between chaffering Ovia to school and amma to shopping, working on MS projects and course work, I rarely find time to paint. But when I do, it becomes my world!!! Hope I would soon post many more of my work :)

How to send an email using mutt in ssh session in a Linux machine?

The process is simple.

1. install a ssh client in your machine. In a command line,
"sudo apt-get install openssh-client openssh-server"

Note: server is optional.

2. You will need a shell account. Its available for free from sdf.org. Just follow the instructions printed on the applet. After creating an account, logout from that account.

3. In the command line, type

"ssh @sdf.lonestar.org"

4. The connection will prompt for the password that you configured with the sdf account. Type it in and press enter.

5. If the password is right, you will have to configure your backspace key and then a crescent will appear and you can start typing your unix commands.

6. For sending mail, you need to open a mail client. I am using mutt. In the command line, type

"mutt"

It will open the inbox. The set of controls that are available will be listed on the top. You will have to setup the /Mail folder, just enter yes and the OS will take care of it.

7. To send a new mail, press "m" or to reply to a mail press "r". It will prompt for To: address and then Subject: After the subject, the pico text editor(may be a different editor) opens and you can type in anything you want. Then press "ctrl+o" to write it out and press "ctrl+x" to exit the editor. The screen will return to the compose mail page. Press "y" to send the message out.

You can explore about other options like adding cc, attaching a file, replying to a group, deleting the mail as you go.

Note: These steps are based on Ubuntu 9.10 OS. If there is a better option to anything I listed here, please share it in the comments field. Thanks in advance.

String reverse in Java

In SPOJ, there is a problem to take two integers and reverse them and find their sum and reverse the sum.

The catch here is when two large integers are given. In java there is a object to help you with reverse, for deleting a char at a index in a string and lot more. It is StringBuffer class.

But the solutions to the problems in C takes very less time and space than Java does. I am not sure how thats possible.

Factorial and their trailing zeroes

In SPOJ, there is a problem to find the number of trailing zeroes in the factorial of a number. Though the question seems intimidating, the solution is rather simple.

zeroes come from multiplying the number by 10. To find the number of times the factorial has been multiplied by 10, we can find the number of times 10 is a factor in the factorial expansion of the number. 5 x 2 = 10 is also possible. So instead of finding the factors of 10, we can find it for 5. All the powers of 5 like, 25,125,etc should also be considered to add number of 5 factors to the expansion.

Thus, to find number of trailing zeroes, keep adding the floor of (num/power of 5) until the division is less than 1. The reason why we ignore the reminders in the factors are we need only whole factors not partial ones.

For more detailed explanation see the Source

BitInteger Java solution for finding factorial

In SPOJ, there is a problem to find factorial for numbers between 1 and 100. Though the algorithm for the problem is very easy and simple, finding the right datatype to handle the factorial of larger numbers was not so easy. But the solution is however made simple by java.math.Integer

Amazing thing about BigInteger, the number of digits that it can hold is limited by the memory of the system. It would be unfair to the rest of people who are trying to find the solution if I post the solution here but I can sure help others and save the time I spent researching on this problem.

Follow this sample program: http://leepoint.net/notes-java/data/numbers/60factorial.html