Livecipher

Car for Sale, Maruti Swift VDI

2011-10-09T17:37:00.002+05:30

Update: The car has been sold out!

I'm selling my car. Details as follows:

*Vehicle Information*
Make	Maruti Swift	Model	VDI (Diesel)
Color	Silky silver	Year	Oct 2008
Kilometers	52500	Asking Rate	-------
*Fuel Economy*
Mileage in Highway	22 kmpl	Mileage in City	17 kmpl
*Features*
Air conditioner	Yes	Leather Seat	Yes
Power Window	Yes	Central Locking	Yes
Child safety lock	Yes	Audio. USB/CD/AUX/FM (Sony)	Yes
Car cover	Yes	Anti Braking system (ABS)	No
Driver airbag	No	Alloy wheel	No

Update: The car has been sold out!

Newto Perforce Guide for Subversion users

2011-09-24T18:23:00.006+05:30

vs.

This blog is intended for developers who are new to Perforce version control system, but used Subversion in the past. It gives an hint of frequently used Perforce commands and equivalent Subversion commands. As you used Subversion in the past, you could understand the Perforce commands quickly and become comfortable with this version control system.

The blog also talks about few features which don't exist in Subversion version control system, yet.

Perforce and Subversion mapping

In following table, we list frequently used Perforce commands and equivalent commands in Subversion.

Perforce Command(s)	Subversion Command	Description
p4 sync	svn checkout	Download the code from the Depot. In Perforce, we call it depot whereas in Subversion we call it repository. In Perforce, we call it workspace, whereas in Subversion we call it working copy.
p4 sync	svn update	Yes! We use the same command p4 sync to upload the workspace. In other words, we use the same command to update the delta, periodically.
p4 edit p4 submit	svn commit	By default, all files in the workspace is read-only. To modify a file, we should mark it for editing. Once it's opened for editing, the write permission is granted. We can modify the file using our favorite editor. Once we are ready to submit, we can use p4 submit command. The changes will be submitted in the Depot and the revision will be generated. We can refer to the change using global revision or running revisions specific to the file.
p4 label p4 tag	svn cp	Creating a branch or tag in Perforce is not cheap! We should create the label using p4 label command. Then, we should associate directories or files to the label. The label thus created will contain the associated entries. When the code is sync'ed using this label, those directories and files will be downloaded or updated, accordingly.

User Authentication

Unlike Subversion, the user authentication is not done during the operation. We can not specify the username and password while doing the changes. Instead, we should login in advance to the depot using p4 login command. The expiry time is configurable at the server side. We should login again if authentication is expired.

GUI Interface

It's a huge boost for Perforce. It has pretty cool user interface for managing the workspace. p4v is popular GUI desktop interface to manage the workspace. It is available in all platforms, Windows, Linux, Unix & Mac. Installation is simple. It's a single executable. We should download it and launch to use it. Simple!

Perforce also has a web based GUI interface, p4web. Using this tool, we could manage the workspace using any web browser. It is used to host a multi-user access to Perforce Depot. Installation is again simple. It's a single executable. It also include a in-built web server. When we launch it, it listens on specific port number, 6000, ~~I think~~. We could access the Depot files by pointing the web browse to this port number. Important aspect of this tool is, it's used for both read & write operations. We could use it to submit changes the Depot, create labels in the Depot, etc.

Client Configuration

The p4 client is a unique interface in Perforce. Using this command, we could configure the mapping for depot and workspace(s). In Subversion, we should get into the working copy directory which we want to update. In Perforce, we could sync the delta using p4 sync command, from any directory. The mapping is read from the setting we do using p4 client command.

Commit Emails

This is another feature I like the most in Perforce. To receive email notifications for any change in depot, we could configure the list of directories using p4 user command. Once it's configured, email notification is sent for any change in the configured directories. In other words, the commit email notification is configured by the user. Unlike Subversion, we don't configure the pre-commit hook script to configure the email addresses.

I'm still a newbie for Perforce. Once I figure out the use case for reintegrate and merge commands, I may record it here. Stay tuned!

Twiki application - 2 cool features

2011-09-21T11:16:00.004+05:30

Twiki application
I have used several wiki applications in the past. It include, Mediawiki, Moinmoin & Jspwiki. Recently I'm using the Twiki application. This post is about 2 cool features in Twiki, I like the most.

Sub-pages

We could embed a wiki page within another wiki page. For instance, we could have a release dashboard. The release may consist of several components. Different team may own each component. Each team may maintain a wiki page to track updates specific to the component. The component page may be embedded in release dashboard page.

This feature is available in-built with Twiki application. Assuming the release name is spartans and you belong to DeploymentAutomation team. You could include the following code snippet in the spartans page.

%INCLUDE{"Spartans.DeploymentAutomation"}%

Thats it! The changes we do in DeploymentAutomation page will appear in the release page, Spartans.

Calender extension

It's available as a plugin. With this plugin, we could add a calendar within wiki page easily. This is highly helpful to integrate Outlook calendar and Team vacation plans in a wiki page. Real collaboration!

The following code snippet is used to add current month and next month.

---++Current Month
%CALENDAR{topic="IndiaCalender" showweekdayheaders="1" todaycolor="aqua" %cellalignment="left" width="80%"}%
---++Next Month
%CALENDAR{topic="IndiaCalender" month="+1" showweekdayheaders="1" %cellalignment="left" width="80%"}%

The nature of the attribute is self explanatory. For instance, The attribute topic specifies the page in which the page the team calendar is available. It could be pointing to the same page. The attribute todaycolor is to specify the background color for today. The above code snippet would display the calendar as follows.

To add a vacation for specific individual, we would specify the dates in the page referenced by topic attribute. The dates can be specified in following format.

* dd Mon YYYY - dd Mon YYYY - Engineer Name

Example:

* 22 Sep 2011 - 23 Sep 2011 - Bhuvan Arumugam

The engineer name(s) are displayed against the corresponding day in the calendar. These features don't seem to exist in other wiki application. If they exist, please let me know.

VPN connectivity using BSNL Dataone

2011-06-25T22:25:00.001+05:30

I'm using BSNL dataone brandband connection. I've been facing issue with accessing websites with VPN. If I do not use VPN, all websites are working fine. I can post/get data from all websites. If I use VPN, I'm unable to post data to any websites. For instance, I'm unable to login to GMail. I'm facing this issue in last one month. I have not faced this issue earlier.

This blog specifies an option to use in your VPN client to get rid of this problem. With this setting, you can post/submit data to all websites.

Operating System

Name: Ubuntu Linux
Version: Maverick 10.04

In Linux, we use the command vpnc to browse websites using VPN. The settings are specified in /etc/vpnc/default.conf file.

bhuvan@rightchoice:/etc$ sudo cat /etc/vpnc/default.conf
IPSec gateway vpn.server.com
IPSec ID vpn
IPSec secret common_password
Xauth username username

Natt Traversal mode

In vpnc, the default natt mode is natt. To fix this issue, we should specify the mode as cisco-udp. We can specify this setting using the following option when we invoke vpnc command.

sudo vpnc --natt-mode cisco-udp

Alternatively, we can specify this setting in /etc/vpnc/default.conf file, as follows:

NAT Traversal Mode cisco-udp

With this setting, we are able to post/submit data to any websites, successfully!

Using Macro with Vim Editor

2010-10-28T12:31:00.001+05:30

This blog talks about howto use macros in Vim editor. The macros in Vim are useful to record series of keystrokes and replay it in other files. The same operation can be performed by writing a shell or perl script. Doing the same with Vim macro is faster.

Let us assume you have around 100 html source file. In all these files, you should replace the following code snippet:

<script language='JavaScript' type='text/javascript' src='http://www.website.com/adserver/adx.js'></script>

<script language='JavaScript' type='text/javascript'>

</script><noscript><a href='http://www.website.com/adserver/adclick.php?n=a60cde47' target='blank'><img src='http://www.website.com/adserver/adview.php?what=zone:9&n=a60cde47' border='0' alt=''></a></noscript>

... with the following line.

<a href='http://www.website.com' target='blank'><img src='http://www.website.com/logo.gif' border='0' alt=''></a>

With Vim, you can do this change in all files in >30mins. The time taken is proportional to number of files. If you have over 100 files and if you have less than 30 mins, you should write a script to do this change.

The logic to do the replacement is, to search for the word adserver, select the line in the visual block, repeat the search for the same word, twice. The 3rd occurrence is the end of the block. Delete the selected visual block. Paste the line to be replaced from a register. To know how to use registers for copying, refer to this link in vim manual.

Pre-requisite: You should have copied the new line, or block of lines in a register. In our example, we use register b. The keystrokes to do this change and record it in a macro is as follows:

qanVnnd"bpqq

The above commands does the desired replacement and record the keystrokes in macro named a.

qa -- Start to record the keystroke. You will see recording text appear in the last line.
n -- Search for the pattern that was previously searched. In this case, it is adserver.
V -- Start the visual selection.
n -- Continue the search. This is the 2nd occurance of the word adserver.
n -- Continue the search. This is the 3rd occurance of the word adserver. This is the end of the selection as well.
d -- Delete the selected visual block. This should delete all the lines mentioned in our example.
"bp -- Paste the content from the vim register b. It is assumed that the line to be pasted is copied in this register before the recording was started. You may use any character as a register.
qq -- End of recording.

Now to replay the macro, you should type the command @a in the file in which you want to do the replacement. Ideally, you may want to open all files in Vim, navigate to all files one by one and type the command @a to execute the macro.
If you are new to Vim, this may sound horrible. But believe me, once you get used to Vim you will find ways to do complex tasks, cheaply!

Release Engineer Opening at CollabNet

2010-10-19T17:28:00.000+05:30

Company: CollabNet
Location: Chennai
Team: Release Engineering

We are looking for a highly talented Release Engineer. The release engineer is responsible to manage the build infrastruture at CollabNet. He will have to deal with build infrastruture for several products/releases on different architecture at a time. He should also manage the installer system for the application on different platform, including RHEL, SUSE, CentOS and MS Windows.

Responsibilities

write scripts to automate build/install process for different products
setup and manage the CI engine (Hudson, Cruise control)
troubleshoot issues with install & build processes
work closely with developers to fix build failures, or propose patches
work closely with QA team to test installer automation process
review the source code committed by your buddies
enhance/fix issues in package manager

Requirements

6+ years of experience. It doesn't really matter if you have strong exposure to the following tools.
proficiency in atleast one scripting languages: Shell/Perl/Python/Ruby
awareness to CI engine: Hudson/Cruisecontrol
awareness to Agile/Scrum/XP software development practices
ability to understand the Java code
awareness to build tools like ant/maven
ability to configure application services like tomcat/jboss
exposure to managing packages using tools like yum/apt-get
well versed in using version control system like Subversion/Git

awareness to advanced topics like merge tracking, would be a plus

ability to prepare rpm spec file and create rpms

if you own a website and/or blog, that would be a plus.

If you believe you are a right fit, please forward your resume to me. Please mention Ref: Livecipher Blog in the email subject.

Wordpress plugin development with multi-site

2010-10-14T17:42:00.000+05:30

If you are familiar with Wordpress, you would know that it supports Multi-site option starting version 3.0. The multi-site is a option to manage multiple websites using single Wordpress instance. The websites may be different from each other by sub domain name or the path.

Activate plugin for all websites

The plugins can be installed at the main website. The plugins can be activated for the main site alone, or for all the websites using Network Activate option in the plugins page.

This article talks about wordpress plugin development with multi-site. Primarily it talks about the issue I came across while attempting to update the plugin setting in the database.

Problem

I had a input form to collect plugin specific options and store in database. The options are stored in wp_blog_options table, where wp_blog_ is the prefix we use for the Wordpress database. For every additional site, this table will be created with blog id prefix. Ex: wp_blog_2_options. When the plugin settings are entered for a particular site, it should add records in this table. But it doesn't happen so.

Solution

I was using the path based setting to access each site. That said, the plugin setting page is different for each site. I was using the following action in the form:

<form method="post" action="<?php echo $_SERVER['PHP_SELF'] . '?page=' . basename(__FILE__); ?>&updated=true">

With this action, the form is submitted to the main site, always. The main site is the one that ends with '/'. This is irrespective of the site I was configuring the plugin for. I figured out the problem is not with the wordpress multi-site setup, but with a bug in my plugin code. I had to rewrite my form with default action.

<form method="post">

With this, the form is submitted to the site specific page and records are inserted in site specific table

Use 2048bit server certificate with Apache

2010-09-21T13:00:00.001+05:30

Lately, I happen to work with Apache most of my time. Here's another finding on using 2048 bit server certificate with Apache. If you refer to Apache FAQ item, it claims that you can't use it:

Why does my 2048-bit private key not work?
The private key sizes for SSL must be either 512 or 1024 bits, for compatibility with certain web browsers. A keysize of 1024 bits is recommended because keys larger than 1024 bits are incompatible with some versions of Netscape Navigator and Microsoft Internet Explorer, and with other browsers that use RSA's BSAFE cryptography toolkit.

It is indeed inaccurate. One of our client wanted to use 2048 bit server certificate. We were doing an R&D on whether it's supported, and if it is not supported, how to bring in the support. When we noticed this FAQ item, we believed that it's not supported. However, when we configured a 2048 bit self signed certificate with Apache 2.2.11, we found it to be working, out-of-box.

The Apache documentation team apparently fixed it in r999259. Refer to the mailing list thread for details.

Tip: If the browser name, Netscape Navigator is mentioned in any Apache documentation page, it is evident that the document is not updated for over 5 years. You may not want to buy the fact, unless you try it yourself.

Proxy to an HTTPS website from a HTTP website

2010-09-08T16:47:00.003+05:30

In my last blog, we covered howto setup proxy to a HTTP website on Debian/Ubuntu systems. Now, we'll cover howto setup proxy to a HTTPS website.

To setup proxy to a HTTPS website, we should enable ssl apache module. This is required even if ssl is not required to render pages in the main site. Execute the following command to enable ssl in Debian/Ubuntu systems.

$ sudo a2enmod ssl

To setup the proxy from http://bestmate.maa.corp.collab.net/release to https://coolmate.maa.corp.collab.net/release, you should add following rules in /etc/apache2/sites-enabled/000-default file in bestmate server:

SSLProxyEngine on
ProxyPass /release https://coolmate.maa.corp.collab.net/release
ProxyPassReverse /release https://coolmate.maa.corp.collab.net/release

Make sure the remote server coolmate is listening on port 443 (https). If your application use any css, js, images you should setup proxy for rendering those pages. You should copy the last 2 lines and replace release with appropriate directory name. For example, to setup proxy for javascripts, assuming they are stored in /js directory, you should add following lines in the same file.

ProxyPass /js https://coolmate.maa.corp.collab.net/js
ProxyPassReverse /js https://coolmate.maa.corp.collab.net/js

Finally, don't forget to restart your httpd server, atleast a graceful restart is required. Wish it saves another few hours for you!

Using mod_proxy on Debian/Ubuntu system

2010-09-07T18:14:00.001+05:30

I came across this unique behavior while setting up mod_proxy in my Ubuntu machine. This behavior is found in Debian systems as well.

We use mod_proxy apache module to redirect requests from one machine to another, without affecting the url in the browser address bar. Consider this use case, a PHP application running in a server named rainbow.maa.corp.collab.net, behind the firewall. The PHP application should be made accessible to the users over internet. For corporate security reasons, the Administrator can not expose the server to the public. This issue can be solved by setting up an Apache server, say drizzle.maa.corp.collab.net, that is accessible over the internet. In this machine, we can setup proxy to server rainbow that is running PHP application. The user may access the PHP application through drizzle, without knowing that it is being rendered from a machine named rainbow.

To make this proxy setup, we should modify the apache configuration in the server, drizzle. No change is required in the server, rainbow. The apache module mod_proxy should be enabled in the server, drizzle. On Debian and Ubuntu systems, you may use a2enmod command to enable the module.

$ sudo a2enmod proxy

This will add following files in /etc/apache2/mods-enabled/ directory.

proxy.conf
proxy.load

The following 2 lines can be added in /etc/apache2/sites-enabled/000-default file to setup the above described proxy system.

ProxyPass /app http://rainbow.maa.corp.collab.net/app
ProxyPassReverse /app http://rainbow.maa.corp.collab.net/app

The ProxyPass directive allows remote server rainbow to be mapped to local server drizzle. It includes all requests sent to remote server rainbow and also the responses sent from it. The ProxyPassReverse directive allows the follow through the redirects, if any.

This setup will be functional only if following line has been added to the apache configuration file, either in the main one, /etc/apache2/apache2.conf or the module specific configuration file, /etc/apache2/mods-enabled/proxy.conf. By default, it's added in file, /etc/apache2/mods-enabled/proxy.conf.

ProxyRequests Off

This is the typical configuration that should work in all systems. However, it doesn't work with Debian and Ubuntu systems. To be specific, it does not work in Debian etch and Ubuntu 10.04 code named Lucid Lynx. We face the following error in apache error log, while accessing the URL: http://drizzle.maa.corp.collab.net/app. The user sees 404 access forbidden error in his browser.

[Tue Sep 07 15:51:34 2010] [error] [client 127.0.0.1] client denied by server configuration: proxy:http://rainbow.maa.corp.collab.net/app

By default, the proxy through localhost is prohibited! A snippet from apache configuration file, /etc/apache2/mods-enabled/proxy.conf.

      AddDefaultCharset off
      Order deny,allow
      Deny from all
      # Allow from localhost

We should uncomment the line # Allow from localhost to allow proxy through localhost. After the change, the snippet may look like:

      AddDefaultCharset off
      Order deny,allow
      Deny from all
      Allow from localhost

After this change, a graceful restart of apache may fix the issue. Wish it saves couple of hours for you!

Hudson Master Slave setup

2010-02-16T16:55:00.006+05:30

Hudson Master Slave setup

(series 2 of 2)

In the previous article, we discussed about the usefulness of the Hudson CI engine. In that article, we discuss about how to setup Hudson and configure jobs to perform various tasks automatically, or on-demand.

The tasks can be scheduled to run on the same machine (Master), or on a different machine (Slave). A master is a installation of Hudson, that can manage one or more slaves. The role of master remains same in Master slave setup. It serves HTTP requests, and it can still build projects on it's own.

Slaves are computers that are setup to build projects triggered from the master. A separate program called slave agent runs on slave computer. In this article we'll discuss about how to setup Hudson to executed distributed builds using Master slave. One computer can be configured to execute multiple slave agents.

How does this work

When the slaves are registered to a master, the master starts distributing loads to slaves. The delegation depends on the specific job. The job can be configured to either execute on the master, or it can be tied to a specific slave. On the other hand, the jobs can be configured to freely roam between slaves, wherein the job is executed using the free slave. As per the user is concerned, the setup is transparent. The results for all jobs are viewable using the Master, irrespective of the Slave that executed the job.

The slave may be built using any Operating system. The Master slave setup is highly helpful while the user has to execute the job on different Operating system. Consider this use case: The application is expected to run on different operating system, Linux, Solaris and Windows system. To address this need, the user can install Hudson on any machine, say Linux, and add 2 slaves: Solaris and Windows. The user can add 3 different jobs, one running on Master itself and others running on slave machines.

Methods to Launch Slave agents

The slave can be launched from Master using any of following methods:

Launch Slave agents on Unix machines via SSH.
Launch Slave agents via JNLP.
Launch Slave agents via execution of command on the master.
Let Hudson control windows slave as a windows service.

The following screenshot illustrates the list of modes under which the Hudson slave can be launched.

Launch Slave agents on Unix machines via SSH

Hudson has a built-in SSH client implementation that it can use to talk to remote sshd and start a slave agent. This is the most convenient and preferred method for Unix slaves, which normally has sshd out-of-the-box. Click Manage Hudson, then Manage Nodes, then click New Node. In this set up, the connection information is supplied, including SSH host name, user name and authentication credentials. The authentication is performed using password or ssh keys. If it is configured to use ssh keys, the SSH public key should be copied to ~/.ssh/authorized_keys file. Hudson will do the rest of the work by itself, including copying the binary needed for a slave agent, and starting/stopping slaves.

Depending on the project and hardware resource availability, the user Desktop can be used as one of Slave, without affecting his day-to-day activities, thus avoiding the need for dedicated Slaves.

Launch slave agent via Java Web Start

Another way of launching slave is to start a slave agent through Java Web Start (JNLP). In this approach, the user will login to the slave node, open a browser and open the slave page using the URL pointed to the Master. It may look like the following URL:

http://masterserver:port/hudson/jnlpJars/slave.jar

The user is presented with the JNLP launch icon. If user click the icon, the Java Web Start kicks in and it launches a slave agent on this computer.

This mode is convenient when the master cannot initiate a connection to slaves, such as when it runs outside a firewall while the rest of the slaves are in the firewall. The disadvantage is, if the machine with a slave agent goes down, the master has no way of re-launching it on its own.

Launch slave agent headlessly

This launch mode uses a mechanism very similar to Java Web Start, except that it runs without using GUI, making it convenient for an execution as a daemon on Unix. To do this, the user should configure this slave to be a JNLP slave by downloading slave.jar, and then from the slave, run a command like this:

java -jar slave.jar -jnlpUrl http://yourserver:port/computer/slave-name/slave-agent.jnlp

The slave.jar file is downloaded from the above mentioned URL. Make sure to replace slave-name with the name of the slave setup in Master.

By default, Hudson runs on port 8080. It can be installed and managed without the need for super user privilege. The super user privilege is not required to manage both Master and Slave.

The following diagram illustrates the list of configuration parameters specific to a slave.

Launch Slave Agent using own scripts

If the above modes is not flexible, the user can write his own script to launch the Slave agent. The script is placed in the Master computer and Hudson runs this script whenever it should connect to the slave. The script may use the remote login program like SSH, RSH to establish connection between Master and slave.

The script would execute the slave agent program like java -jar slave.jar. The stdin and stdout for the script should be connected to the master. For example, the script that does ssh myslave java -jar ~/bin/slave.jar would satisfy this need, when it is executed from the Master web interface. For this reason, running this script manually from the command line does no good.

The copy of slave.jar can be downloaded from the above mentioned URL. Launching the slave agent using this mode requires additional setup in the Slave. The benefit is that when the connection goes bad, the user can use Hudson web interface to re-establish the connection.

The user should update slave.jar every time the Hudson system is upgraded. Though, in practice, the slave.jar is not changed frequently.

About the Author

By: Nithya Pattabiraman. The author is currently employed in CollabNet working for their Quality Assurance Team. She can be contacted at nithya AT collab.net.

Six traits of a Good Leader

2010-01-27T11:36:00.005+05:30

APJ's motivational interview with Knowledge@Wharton. Courtesy: Found this link from Sriram's Video Collection.

Few abstracts from the interview.

What's Creativity?

Learning gives Creativity;
Creativity leads to Thinking;
Thinking provides Knowledge;
Knowledge makes you great!

Six traits of a Good leader:

A leader should have a vision
A leader should travel in unexplored path
A leader should know how to manage Success & Failures
A leader should have courage to take decisions
A leader should have a nobility in Management
Every action of a leader should be transparent

Agile Methodology: Continuos Integration with Hudson

2010-01-20T14:38:00.002+05:30

Hudson - an extensible Continuous Integration engine

(series 1 of 2)

If you believe in Agile software development practices, you already realize the importance of continuous integration. It is one of Agile methods to facilitate the delivery of bug free software to the customer. The following practices define the true continuous integration system.

Members integrate source code frequently
Each integration is verified by automatic build
The code is tested in a mock production environment

The effects of using the continuous integration system are:

Integration issues are found early
The status of latest build are easily known
The current build is available instantly for demo or release purposes

The Figure 1 illustrates the architecture of a Continuous Integration System.

Why Hudson?

The Hudson system is designed precisely to meet the continuous integration needs. It is useful to build and test software projects continuously. It provides an easy to use interface,, making it easier for developers to integrate changes to the project. The automated and continuous build increases the productivity of the team. It is also useful to monitor the externally executed builds and notify the interested parties when something is wrong. The Developers can save the changes in the Version control system. The Hudson job(s) monitor the VCS continuously and act upon accordingly.

The Figure 2 is a sceenshot from Apache build farm that gives an hint on how the Hudson home page would look like when you configure multiple jobs.

Features

Easy installation. No addition installation instructions, no database. java -jar hudson.war
Java and Open Source. The Hudson system is fully written in Java language, thus it is platform independent. The Hudson system is Open source, released under MIT License.
Easy configuration. The Hudson system can be configured completely using the web GUI. There is no need to tweak or configure the settings manually.
Distributed builds. Hudson can distribute build/test loads to multiple computers, thus utilizing the hardware resources effectively.
Junit/TestNG test reporting. The Hudson system can parse the Junit/TestNG test output generated by build/test scripts and generate the intuitive report.
Plugins. The functionality is extended using various plugins. You can write own plugins to support tools and processes you use. You can visit the following link to view all available Hudson plugins.
http://wiki.hudson-ci.org/display/HUDSON/Plugins

Jobs

The build in Hudson is configured as a Job. It may involve one or many tasks including, code compilation, executing unit tests, publishing build artifacts, updating issue trackers, identifying code violations, generating documentation, etc. The build can be configured as a series of shell, perl and python commands. The elegant method is to write a script using any scripting language that perform all activities and execute the script from a Hudson job. Depending on the plugins installed, the user may see many options in the Job configuration screen. The job may belong to any of following category:

Scheduled. The Job can be scheduled to run at specific date/time. The Linux crontab like scheduling is possible.
Monitor Version Control System (VCS). The Job can be configured to run immediate after a commit in the VCS. The Hudson system can poll the VCS periodically, detect the changes if any and initiate the build process.
On demand. The build can be triggered on a demand basis.

In short, the job can be configured to update the VCS working copy, execute the build, generate the installation packages, run tests on different platform, generate documentation and code quality reports. The final result is published in the user interface.

Plugins

The plugins can be managed from within the Hudson user interface. The users has the ability to keep track of installed plugins and new plugins. The updates to Hudson and plugins are tracked from the same screen. The users are notified of updates to Hudson and plugins (if any) automatically.

Few cool plugins :

Leader board
Plugin to report code violation
Webtest plugin
Publish Documents
Integration with ftp/scp servers
Integration with Issue trackers

Conclusion

Hudson home page: http://hudson-ci.org/

With the above suggestions, you may configure a simple Hudson Job that performs certain tasks specific to build and test. Once you start using Hudson system, you will be interested to explore the scope of using Hudson system to execute distributed builds. The Hudson system can be configured to run multiple Slave instances in parallel on different platform. We will cover the Hudson Master Slave setup in next series. Stay tuned!