Old Machines Lying around? Pic courtesy: HowStuffWorks.com

If you’ve got some old legacy hardware catching dust at your work place, you might finally be able to get some functionality out of them, thanks to cluster computing techniques. In this part, we show you how to create a simple MySQL Cluster that you can use for a variety of database related applications.

What you need:

• At least three fully functional machines with an NIC and minimum of 40 gigabytes of HDD space (May need more, depending on the data sizes and the level of redundancy needed)
• Connectivity cables (Cat-5 Ethernet cables with RJ45 Connectors on both ends)
• One Ethernet Switch (a hub will also do, but a switch is considered to be better)
• Ubuntu (or another GNU/Linux Operating System). Get it Here
• The MySQL Cluster Package: This is not the same as the MySQL package that is commonly used. Get it Here
• Basic GNU/Linux Networking Skills. Get them Here
• Patience, Control over your senses and Calmness of Mind. Get it Here

What you will build:

• An SQL Node(front-end): These are the nodes that will receive the database queries from the user, and pass on the necessary details to the data nodes.
• A Data Node (back-end): These nodes are the ones that actually hold the data, and this is where all the manipulation will actually take place.
• A Management Node: Think of this as a Master node, that coordinates the data and SQL nodes, and ensures that they are all active and functioning.

Configuration Details

To properly build a cluster, the network must be properly set up. Here are the IP configuration details that you need to set up with the three machines. Note that the Hostnames may be changed to whatever you like.

Management Node:

• hostname: Venom,
• private IP: 192.168.5.1,
• tcp port: 2205

SQL Node:

• hostname: Eris,
• private IP: 192.168.5.2

Data Node:

• hostname: Olympia,
• private IP: 192.168.5.4

Installation commands on All Three Machines

The following steps must be performed on all three machines:

• Create the MySQL user with:    adduser mysql
• Create the MySQL group with:  groupadd mysql
• Extract the downloaded package in /usr/local directory with:   tar zxvf ~/mysql-cluster-gpl-7.1.14-linux-i686-glibc23.tar.gz
• Create Symbolic Links with:   ln -s mysql-cluster-gpl-7.1.14-linux-i686-glibc23 mysql-cluster
• Change to the directory with:   cd mysql-cluster
• Change Ownership to the MySQL user with:   chown -R mysql:mysql /usr/local/mysql-cluster/

Installation and Activation commands on the Management Node

• Create Directories with: 
• mkdir -p /var/lib/mysql-mgmd-config-cache
•  mkdir -p /var/lib/mysql-mgmd-data
• Add the following by editing the config.ini file with:   nano /var/lib/mysql-mgmd-data/config.ini[ndbd default]NoOfReplicas = 1
  DataDir= /var/lib/mysql-ndb-data
  DataMemory = 64M
  IndexMemory = 128M[ndb_mgmd]
  # Management process options:
  DataDir = /var/lib/mysql-mgmd-data
  PortNumber = 2205
  HostName = [ndbd]
  # Options for data node ndb1
  hostname =
  
  [mysqld]
  # SQL node options for dbsrv1
  hostname = 

• Add the IP addresses in each case, as chosen.

Installation and Activation commands on the Data Node

• Run the Install Script with:   scripts/mysql_install_db –user=mysql –datadir=/var/lib/mysql-ndb-data

Installation and Activation commands on the SQL Node

• Install a dependency with:   apt-get install libaio1  (command may differ for other OSes)
• Run the install script with:  scripts/mysql_install_db –user=mysql –datadir=/var/lib/mysql-node-data –basedir=/usr/local/mysql-cluster
• Edit the /etc/mysqld-cluster.cf file and add the following:
  [mysqld]
  ndbcluster
  ndb-connectstring=192.168.0.100:2205

Starting the Cluster

• In the Management Node
• Change to the bin directory with:    cd /usr/local/mysql-cluster/bin
• Start the Management Node with: ./ndb_mgmd –initial –configdir=/var/lib/mysql-mgmd-config-cache –config-file=/var/lib/mysql-mgmd-data/config.ini
• In the SQL Node
• Start the SQL Node with:    bin/mysqld_safe –defaults-extra-file=/etc/mysqld-cluster.cf –user=mysql –datadir=/var/lib/mysql-node-data –basedir=/usr/local/mysql-cluster &
• In the Data Node
• Start the data node with:   bin/ndbd –ndb-connectstring=192.168.0.100:2205

Thats It!

The Cluster must now be up and running! Point the clients to access  mysql on the sql  node, and you are good to go!

We live in a world that demands high performance, in almost every aspect of life. With the advent of high performance computing, we have reached a level where machines with higher and higher configurations are being built in order to satisfy the growing demand for power and performance at higher speeds. Before going into the depth of the project itself, there is a need to understand the basic concepts that drive the world of high performance computing.

Today, most organizations, educational institutions and firms invest in their own mainframes, servers or high end machines to run their data centers and perform their personal analytics. However, the fact of the matter is that the resources in such cases are being underutilized, even though steps have been taken to ensure that the mainframes or compute nodes are not left idling. Up to an extent, the level of wastage has been reduced by introducing slot based usage of the high end machines, and by introduction of Clustering algorithms but there is still a lot of wastage despite these measures.

In any real world organizational setting, we often find a consistent cycle of hardware upgrades and we often find older legacy systems catching dust in the background. While reselling these systems can get back a percentage of their value, even the notion of profiting from such a move is unthinkable. These legacy systems often consist of old processors which are not too power hungry, and yet, not too slow either. So why not make use of them when they are still live and kicking?

Setting up a simple Cluster with Ganglia Monitoring

The Ganglia monitoring tools are pretty useful, and provide web based interfaces that can be used to monitor many intimate details of each server and node. The monitoring tool can be set up as follows:

1. Installation on the Monitoring Server:

The dependencies are: Apache2, php, libconfigure, libexpat1-dev, libpcre3-dev, and libapr1-dev. Easiest way to install it is to use binary packages. On Ubuntu/Debian to install, we can use apt-get:

apt-get install ganglia-monitor gmetad libganglia1 ganglia-webfrontend

This installs the monitoring daemon (gmond) and the Meta Daemon (gmetad) as well, along with all it’s dependencies. Depending on the existing firewall rules, we may need to add extra rules. However, we had no special firewalling schemes, as the set up was fresh.

2. Configuration of the Server and Nodes:

To properly configure the cluster monitor, we need to tell the monitor that there are nodes and servers in the same cluster. We used a single cluster configuration set up, but Ganglia is capable of monitoring a multi cluster set up as well. The image given below represents Ganglia’s multi-cluster monitoring capabilities:

Next, we need to create a config file which will describe the cluster every time Ganglia starts up. This is done on all the nodes and the master, by running the command on the server:

gmond –default_config > gmond.conf

This creates a config file called gmond.conf. This will have a default template that can then be altered to suit the tastes of the cluster.

3. Starting Ganglia-monitor

Starting the monitor requires the following steps:

1. Start the master and nodes

2. Run the command to start the monitoring daemon on server:

gmond -c gmond.conf

3. Run the gmetad daemon on server:

gmetad

4. Run the gmond daemon on the nodes :

gmond -c gmond.conf

Now, the monitoring interface can be accessed, by running the web-interface http://localhost/ganglia-web. Note that the part after localhost needs to be what is used to name the front-end directory in the machine.

Here is a Screenshot of a configured Ganglia Front-End

Ganglia Front-End showing the Load Metrics of each node of the cluster

More to Come!

For those who want to know more, the next post will be on setting up an effective cluster of 4 machines, which will all work together as a MySQL Cluster. MySQL clusters are pretty useful specially when made up of older systems, in that they can take higher loads from clients and can make best use of old hardware that would be collecting dust otherwise. This cluster would be ideal for an organisation or educational institution where the client-server architecture is followed. Stay tuned for the next post of this 3 post series, where we will show you how to make the best use of those old machines you may have lying around.

Writing a proposal is one of the hardest things of GSoC, especially if you are new to the organization and project and consequently do not have a clear understanding of the code base and its architecture.

A few things to keep in mind before starting off on writing your proposal -

1. Know Your Mentor/Organization’s Understanding of the Project
The first step before writing a proposal would be to completely understand the project requirements and what the mentor envisions the project as. Many project descriptions are very brief and do not give you the complete picture. You can mail the mentor/organization about your understanding of the project and ask them to correct you if you are wrong. Mentors love to help out people who have done some background work and who don’t ask obvious questions.

2. Revision and Iterations

My proposal went through 4 rounds of evaluation before it was complete. The process of evaluation started a month before the submission date. I had my final proposal ready even before the submissions began.

A proposal for a project is an iteration of changes and improvements. It usually cannot be complete in a single draft. If you are writing a proposal, make sure you do a thorough research on the project, code architecture and requirements. Write a first draft of the proposal quickly and send it to the mailing list of the organization or the mentor and ask him to point out any mistakes. That is why it is important to start early so that you give yourself time for further iterations and improvements. Shed your inhibitions and mail your proposal! The same set of people are going to look at your proposal later on anyway.

3. Mailing Your Proposal
Mailing your proposal to the organization helps everyone in the organization to have a look at it and will fetch you multiple comments and criticisms which are beneficial on the longer run to your final draft.

Note : Some organization recommend not mailing your proposal to public mailing lists to prevent plagiarism. Find out from your mentor how he wants the proposal to be submitted for evaluation.

Here a few selected proposals from yesteryears – http://goo.gl/EvyrX

Outline of a proposal -

Few organizations have their own proposal template, and it is sometimes mandatory that you follow their template. Else, you can come up with your own version. Proposals usually contain these sections -

1. Brief introduction to your project – a 4-5 line introduction to the problem statement.
2. Why is the project needed? – very important as organizations have limited slots for accepting projects.
3. Implementation details – you talk about the architecture and the design you will follow.
4. Timeline – the most important section of the proposal. Here you give them a detailed timeline of how you are going to proceed with the project including the tentative dates.
5. Future improvements – if needed.
6. Your bio – you need to talk about your prior experience with FOSS and contributions. It is ok if you are first timer. GSoC is about getting newbies to contribute to open source anyway. But you stand a better chance if you have contributed to some open source project beforehand.

Details:

We have been working hard on building a new Google Play Developer Console, which we have rolled out to all users. The old version of the Google Play Developer Console is now deprecated. For a short time, you can switch back to the old developer console using the instructions below.

Below are some changes that you’ll see in the new version of the Google Play Developer Console:

• Updated look and feel.
• New publishing experience, such as the ability to compare new and old APK versions within the console.
• Create store listings in more languages.
• Statistics on user ratings and reviews.
• Updating your app in the new site makes automated store listing translations available to users on the web.
• Localized graphics for each of the languages where you provide a translation.

Note that if you are experiencing issues with the new Developer Console, please report them using the Feedback link at the top of the page in the Developer Console. The feedback link can also be found in our help center.

Switching between new and old design (This is Not Possible after Aprill 15th 2013)

Whenever you need to use a feature which is not yet supported in the new Developer Console design, simply click on View in the old site.. To switch back, click on View new design button.

Compare to other places many of the Google servers are a little bit slow in India. But the good news is Google will install more servers in Asia to improve internet speed by 30% great news for developers/users.

More Info: Now is the time to switch to the new Google Play Developer Console | Android Developers Blog:

As we have posted here about India is listed as supported country on Nexus Store, here is the next announcement

Nexus 7 is now officially available in India via Google Play store, Priced at 16,000 Rs and Google says it will be shipping around 5th April 2013.

Hope to see other Nexus devices in India soon.

Hit the link below & enjoy

Nexus 7 (16GB) – Google Play

Wikipedia:

In software development, Git (/ɡɪt/) is a distributed revision control and source code management (SCM) system with an emphasis on speed.^[3] Initially designed and developed by Linus Torvalds for Linux kernel development, Git has since been adopted by many other projects.

Every Git working directory is a full-fledged repository with complete history and full revision tracking capabilities, not dependent on network access or a central server.

Git is free software distributed under the terms of the GNU General Public License version 2.

Git isn’t big deal, but many finds it very tough to begin with. To master git its matter of time. Here comes a simple note on how you can start with git and push your first code to github. Eventually you can grow into git-master. So lets begin!

1) Create A Repo

go to www.github.com and sign up. I leave it to you to install git. Even you haven’t configured your ssh keys, its cool. I will go “http” way in which you ain’t need to manage your ssh keys with github. If you are thinking what the hell is “ssh” keys then just ignore it.

So, by now I assume you have signed up for github and you have logged it.

Lets create a repo from top right corner as shown in image below.

Next Page will look like image below:

Describe your repo, initiate with README.md and your repo is ready

Done. Click “Create repository” and come back to your terminal.

➜  ~  mkdir my_git_repos         (choose a directory name of your choice)
➜  ~  cd my_git_repos

copy the link from your repository, we will need it in a while.

you are inside the folder “my_git_repos”,

2) Clone the Repo

now hit the following:

➜  my_git_repos~  git clone https://github.com/linuxcandy/sample.git
Cloning into ‘sample’…
remote: Counting objects: 3, done.
remote: Total 3 (delta 0), reused 0 (delta 0)
Unpacking objects: 100% (3/3), done.
➜  my_git_repos~  ls
sample
➜  my_git_repos~  cd sample
➜  sample ~  ls
README.md

hit above commands and you should see exactly like what is shown above. Let me quickly tell you about what happened. “mkdir” created a directory, “cd <directory name>” took you into the directory, “git clone” cloned a version of what is their on github.com of your repository to your computer in the folder named “sample” which is your repository name and “cd sample” took you inside your repository on your local machine, and finally “ls” told you that “hey dude, you have a README file in here” . Congratulations!!

Cool, not lets push some code from our local machine to our github repository.
Create a file inside “sample” folder. I have created a “my_code.py” file. and inserted a “hello git world” line in it. I am going to show you how to push this to our repo in github.

From your terminal hit the bold codes below without thinking what they are going to do, and I will tell you after you are don.

3) Commit and Push a Code

➜  sample~ git status
# On branch master
# Untracked files:
#   (use “git add <file>…” to include in what will be committed)
#
#    my_code.py
nothing added to commit but untracked files present (use “git add” to track)
➜  sample~ git add my_code.py
➜  sample~ git commit -m “my first code to git”
[master 0780b5f] my first code to git
1 file changed, 1 insertion(+)
create mode 100644 my_code.py
➜  sample~ git push
Username for ‘https://github.com’: linuxcandy
Password for ‘https://linuxcandy@github.com’:
Counting objects: 4, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (2/2), done.
Writing objects: 100% (3/3), 315 bytes, done.
Total 3 (delta 0), reused 0 (delta 0)
To https://github.com/linuxcandy/sample.git
83c5881..0780b5f  master -> master
➜  sample~ git status
# On branch master
nothing to commit (working directory clean)
➜  sample~ cat my_code.py
print “hello git world”

Congratulations again! you have pushed your code to your repo in github. Don’t believe it ? Lets have a look.

Go to your repositry on github and hit refresh.

You see ? you have your code there!
ok now let you tell you what you did or what the codes above do.

git status -> to check which stage you are in. it will show you a detail about your status.

git add <filename> ->  after you make changes in a file. you need to add it

git commit -m “<message>” -> this commits your addition of file above with a short message about your edit.

git push -> pushes your commits, changes to repo remotely. Here you are done!

Wait, dont rush. there is one thing I forget to tel you blame my memory . Here we created our own repo, but what if we want to work on somebody else’s repo ? or want to clone it and start contributing or helping a friend in his/her repo at github ?

Everything is same like above but before you push any code you need to “pull” latest code because this time you are not the only one which will be working on the code. Working in a team on a same code is different.

so you need to follow like this:

Pull > Make Changes > Add <edited file names> > Commit -m “<message>” > Push

Remember the rule, and you are a git hero!

Wanna be super hero ? Read this book http://git-scm.com/book

Adiaux!

Ubuntu Touch has already made lots of news and hence I decided to try it on  my LG Nexus 4. An important thing to note is that the currently released version of Ubuntu Touch is just a developer preview and hence far from ready for the general usage. However it does have enough features to just give a peek on how beautiful a fully functional Ubuntu Touch would be.

The main concept of the Touch UI is that user uses all the edges of the phone. It also depends a lot on swipes and hence requires your screen to be really receptive to touches.

The installation was a breezer with complete detailed information given here, https://wiki.ubuntu.com/Touch/Install. It did take some time to install thanks to my slow broadband connectivity. The boot time is less than 10 secs though currently the start-up looks dull without any animation. First screen shown is a circle which is supposed to contain all the latest updated messages from different application. However it currently doesn’t show anything and is just a placeholder. As you swipe over it, you are shown the home screen. This home screen is supposed to be the snapshot of your complete phone, where it shows the frequently used apps, recently contacted people, recently listened music and so on. As you swipe from right to left, you get dedicated screens for Apps, Music, People, Videos. The main concept of the Touch UI is that user uses all the edges of the phone. It also depends a lot on swipes and hence requires your screen to be really receptive to touches. The left edge panel is dedicated to show all the frequently used apps. Take your thumb to the left edge and just swipe up and down to select any of the frequently used apps. At the bottom of this panel is an option to go back to your home screen. On top left, there is a search option which is supposed to be a universal search across all data in your phone. However in the current developer preview it is not functional yet. The top edge panel is dedicated to the settings of the phone. A swipe over the top edge and you can choose to modify the settings of Battery, Data and time, Networks, Messages, Sound. In short everything to play around with your settings. All the option in the settings are working fine. Some of the basic functionality like autorotation of the screen, disabling wireless are still missing. Due to this, the wireless is always on. The right edge is used to bring the apps running in your background to focus. So shifting between apps is just a swipe from right to left.

Currently there is no way to stop your running apps and hence it means any apps that you have tried will keep running on the background till you restart.Again since this is a developer preview, this is not much of an issue. The bottom edge of the screen is owned by the app and hence this is app specific.

Over all the concept of UI making use of all the edges is very intuitive. However I do need to see how the UI will behave when autorotation is enabled. Currently when I start the music player, autorotation does happen, but the edges do not change their behavior.

From the hardware side, both the cameras are detected and working fine. The flash can be enabled disabled for the back camera. On the video mode, you can enable your flash to be on continuously. This is good to capture video in dark. However I was not able to capture the video itself. I guess, It is not yet available in the current release.

Wifi was detected fine and I have able to browse over internet without any issues. However most of the website detects the phone as an iOS and suggests me to download the iphone app.

Power consumption wise, it seems to be to early to say anything. It did last 24 hours but with very less usage. There is certainly heat generated and you can feel the heat at the back of the phone. This might be due to the continuous apps running on the background.

I was able to successfully make phone calls and send a message from it. The sound quality was good. I tried connecting my earphone but I was not able to hear anything on it. The notification is not very good as I need to go to the call window to find out if I have any miss call or new messages. May be when the first screen becomes active, these notifications will also be shown there. Also the speaker doesn’t seem to be working while talking during a call.

Few areas that I have not tested yet: 1) Audio/Video: Music and Video apps are just snapshots and are not
functional. I was not able to get Youtube to play a video on the browser. Hence overall have not been able to check the audio/video part of the phone other than listening to caller tune and voice over phone. Few points to improve:
1) UI though intuitive still needs sometime to get adjusted to.
2) Startup seems to be dull with just the first circle of latest update seen. Some animations would be nice.

Overall it seems to be a great promise and I am very eager to follow its development. The SDK to develop apps is already released and I am eager to write my first hello world app running on Ubuntu Touch. .

For updated news about the development, please check out this link,

https://wiki.ubuntu.com/Touch/ReleaseNotes

Sundar Pichai Takes Over For Andy Rubin As Head Of Android At Google, Signals The Unification of Android, Chrome And Apps.

Andy Rubin, who built Android into a software platform used by most of the world’s largest handset manufacturers from Samsung Electronics Co Ltd to HTC Corp, will “start a new chapter at Google”,Larry Page  wrote without elaborating.

More Info: Google official Blog

Screenshot taken on 12-03-2013 08:56PM IST

Closeup

Access to the Playstore for Devices is not available for Indian visitor, but we can expect that launching in India soon.

One another possibility can be that this option refers to official launching of Nexus 7 tablets (Earlier available in Croma stores) in Indian market.

Chatwing’s universally free chat box can be installed to any blog or website. Once installed, it can provide a generous stream of traffic to the website. This is useful for any Linux blog that is offering guides, tutorials, walkthroughs, and user interaction. With Chatwing’s chat app, website visitors can post their Linux-related questions and even ask about apprenticeship from experienced developers in the blog.

The Linux jargon may also be confusing to previous Windows users, but Chatwing chat box can remove this problem entirely. If the Linux blog has an active admin, visitors can gain the information they need within seconds. This is one of the strongest factors of Chatwing tool: real-time interaction. Adding to that, Chatwing also offers social media integration. This extends the Linux support twofold. Now, website visitors can join the website owner’s own Facebook page (if he has one) and they can even find new communities suggested by other Chatwing chatters.

Chatwing chat software is also dynamic because it allows multiple customization options for all of its registered users. The color and size of the chatroom can be changed anytime, according to the preferences of the admin. With this simple communication tool, introducing the Linux system to everyone will become a breeze.
You can find Chatwing online http://chatwing.com