About the Author

Latha Annur Subramaniam is a visual thinker, a technology enthusiast and a person with multifarious interests. She was a core engineering member at Selectica, a valley based pioneer in online Guided Selling systems and later at Cognizant’s Digital Security Practice, working for Media and Banking giants. Having garnered the unique combination of Technology, Business and Creativity, she now runs a boutique Marketing Communications Consulting company.

Disclaimer: The logos in this cartoon clip is meant for humourous purposes only. No offense intended.

Starting today till July 1, 2012, Microsoft will receive online applications from startups. Any application that can be built using the Cloud platform and can operate in any business or market segment is eligible to apply. Ten applicants, selected through a rigorous process of screening, will be invited to join the program in Microsoft’s office in downtown Bangalore, starting September 3, 2012. For the next four months, they will have access to the best workspace, finest infrastructure, top mentors and all the tools needed to create the most successful startups.

The selected startups will also have access to all the resources of Microsoft’s BizSpark program and $60,000 in Azure credit through the Microsoft BizSpark Plus program. They will be able to leverage the unlimited possibilities enabled by an open and flexible cloud platform, the Metro user interface on Windows 8 and Windows Phone, and much more.

A key ingredient of this program is the exclusive mentorship and coaching that the startups will receive from successful entrepreneurs and leading Venture Capitalists, and from top technical and business leaders in the industry, as well as at Microsoft.

“The Microsoft Accelerator for Windows Azure is a particularly exciting program and we are thrilled to launch this initiative in India. Through this unique initiative, Microsoft is partnering with leading industry mentors to provide startups with the much needed support they need to innovate and become successful businesses. We are not only providing them technical and business guidance and mentoring, but also connecting them to prospective VCs and investors. The success of this program will be measured by the success of these startups. Microsoft will have no stake in these companies,” said Amit Chatterjee, Managing Director, Microsoft India (R&D) Pvt Ltd.

Microsoft is in the process of finalizing its roster of mentors from across the Indian technology and business landscape as well as its own leadership base. Registration for the accelerator opened today and startups from all over the world are invited to submit their applications here: http://www.microsoft.com/india/accelerator. The list of 10 shortlisted startups will be announced in July and classes will commence in September. At the end of the 4-month period, Microsoft will organize a Demo day where the startups will pitch their ideas to key industry influencers and to top angel investors and VCs for funding.

The fundamental unit of deployment in Windows Azure is a Role. A Role is the template that contains the actual bits and the configuration. It is a mechanism for the developers to deal with the Compute (which is called as a Hosted Service) part of the Azure. New and existing applications targeting the Cloud need to be packaged into one or more Roles. For running web applications, Windows Azure offers a template called WebRole. ASP.NET and PHP applications will be deployed as WebRoles since they are exposed on HTTP 80 or 443 to the outside world as web applications. For running non-UI, long running processes, the code needs to be deployed as a Worker Role. A WorkerRole is designed to run the “raw” code that can be run on Microsoft Windows Server 2008 OS with an exception of UI based applications. Windows Azure also offers another Role called a VMRole that can be used to run applications that do not fit into the Web Role and Worker Role types. VMRole is typically used to run legacy applications and components with dependencies that may not run on the other two roles. An in-depth discussion around Roles is outside the scope of this article. We will focus on the way an application is packaged and the techniques of deploying it on Windows Azure.

Typical Windows Azure application contains three elements – 1) One or more Roles, 2) the service definition of the application and, 3) the service configuration.

1. Role – This contains the ASP.NET web application packaged as a WebRole or code that is packaged into WorkerRole. This contains the usual .NET code and a mandatory class for Role Entry point in the form of a WebRole.cs and WorkerRole.cs. This combination of Role EntryPoint and the traditional code turns the project into a WebRole or a WorkerRole.




2. Service Definition – This XML file (typically called ServiceDefinition.CSDEF) enumerates all the Roles that will be deployed into one Hosted Service. The settings that need to be defined once and before running the instances go into this file. Settings like the size of the VM, the certificates to be included for the Hosting Service, definitions of custom settings that can be passed to the application at runtime, input and internal endpoints for the communication, local storage space per role, and finally the startup tasks that need to be executed. In summary, this file has the global settings for the Hosted service and the settings that are required per role.




3. Service Configuration – This XML file (typically called ServiceConfiguration.XXX.CSCFG where XXX can be Cloud or Local) contains the settings that can be changed without forcing redeployment of the application. For example, the number of instances of a specific role can be changed in this file that would impact your deployment immediately by spinning up more VMs based on a specific role. Though Windows Azure is a PaaS, it gives the ability to choose a Guest OS version for better compatibility of the runtime. Developers can opt for a specific Guest OS that comes with a specific version of .NET runtime, patches and features that map to a specific version of Azure SDK. By changing the setting in the Service Configuration file, developers can switch between a variety of Guest OS releases. Custom settings defined in the Service Definition file can be changed at runtime forcing the application to refresh itself with the new settings.

So, a Windows Azure application contains one or more Roles, Service Definition and Service Configuration.

Once we have all these in place, we need to package the application for the Cloud. In the next step, we will explore the options to package an application.

The packaging process will emit two files that need to be deployed. One of the two files is the CSPKG file that contains the compressed version of the application combined with the settings mentioned in the CSDEF file. The other file is the ServiceConfiguration.CSCFG file that contains the runtime settings that can be changed during the lifecycle of an application. But what are the options to generate these two files? We can use one of the three techniques to complete the packaging.

1. Visual Studio – Once the Windows Azure Tools for Visual Studio are installed, developers can enjoy the single click deployment from the IDE. Visual Studio offers both packaging and publishing of the Cloud applications. This is the most intuitive and productive way of deploying applications to the Cloud.




2. CSPack command line tool – This comes as a part of Windows Azure SDK. CSPack can be used to package applications for both Windows Azure Emulator running locally and Windows Azure.




3. MSBuild – Developers can use MSBuild to prepare an application package for Windows Azure. The MSBuild XML file will have the appropriate tools to invoke the compilation and packaging the application.

All the above tools will emit the final .CSPkg file and .CSCFG file ready to be deployed.

This article looked at what goes into a Cloud application targeting Windows Azure. We discussed the extra elements that needed to be added to conventional .NET applications to transform them into Cloud applications. This is essentially the first step in going to the Cloud. In the next part of this article, we will take a closer look at the deployment and how various tools interact with the Windows Azure control plane. Stay tuned!

- Janakiram MSV, Chief Editor, CloudStory.in

PaaS is fundamentally designed to abstract the nuts and bolts of the underlying Cloud Computing infrastructure. It hides the complexity of Cloud by just exposing what is required for the developers to deploy the code. While developers need to focus on the tools and the IDEs to deal with the PaaS, it is certainly useful to understand the underpinnings of the PaaS.

CloudStory.in brings you a series of articles that goes behind the scenes of a PaaS deployment. Typically developers write code and deploy it on a PaaS that the end users will consume. In between these two stages of deploying the application and consuming the application, a lot of interesting things happen. The objective of this series is to analyze the workflow that takes place between deployment and the availability of the application. Though every implementation is different, there is a common framework that the major PaaS implementations rely on. For example, there is an uncanny resemblance between Windows Azure and VMWare Cloud Foundry. By end of this series, you will be able to relate to the common concepts and terminology across multiple PaaS offerings. We will cover the under the hood details of popular PaaS implementations like Microsoft Windows Azure, AWS Elastic Beanstalk, Cloud Foundry and Red Hat OpenShift. So, fasten your seat belts and get ready for an intriguing ride of PaaS internals.

- Janakiram MSV, Chief Editor, CloudStory.in

About the Author

Latha Annur Subramaniam is a visual thinker, a technology enthusiast and a person with multifarious interests. She was a core engineering member at Selectica, a valley based pioneer in online Guided Selling systems and later at Cognizant’s Digital Security Practice, working for Media and Banking giants. Having garnered the unique combination of Technology, Business and Creativity, she now runs a boutique Marketing Communications Consulting company.

Disclaimer: The logos in this cartoon clip is meant for humourous purposes only. No offense intended.

Founded by CEO Suresh Sambandam and CTO Mani Doraisamy,OrangeScape provides Visual Platform as a Service (PaaS), a modeling driven visual development and deployment environment for creating business applications. OrangeScape is the only cross-cloud provider in the world that supports all major public cloud platforms.

With this funding, OrangeScape plans to expand its services to mature economies likeUSA,UKandWestern Europe. Additionally, the company will focus on US based marketing, sales, customer support and remove any geographical, psychological and enterprise support related barriers that exist in US and Europe.

OrangeScape’s patented rule engine technology can be leveraged by both large enterprises and Independent Software Vendors building SaaS as well as “Enterprise PaaS”. The Visual PaaS platform is particularly suitable for high volume enterprise usage giving a superior price-performance edge.

OrangeScape, that has customers and partners spread across India, USA, UKand Europe, features in the list of Global 10 Cloud PaaS providers, by IT intelligence firms Forrester and Gartner. Forrester expects the PaaS market to reach $11.9 billion by 2020, up significantly from the $820 million in 2011.

Suresh Sambandam said, “Using funds from IAN and the network strength of Members of the Indian Angel Network, we are now set to add global enterprise customers aggressively. With the strength of our technology we will quickly move up in the Global Top 10 Platform as a Service listing.”

OrangeScape’s major enterprise customers include United Biscuits, Unilever, AstraZeneca, Citibank,Fullerton, Geojit, IndiaInfoline, Pfizer and Sterlite.

IAN investors also build a strong foundation for their investee companies, helping them with their key management, introduction of customers, building global footprint, and assist in the next round of investments or acquisition by global players.

For more information, please visit their website and you can also read guest columns by Suresh Sambandam here.

Suresh Sambandam, Founder & CEO at OrangeScape, said on this partnership – “With our Visual Platform as a Service (PaaS) we address business application and workflow needs of very large enterprise customers focused on productivity through platform abstraction. This partnership with Eucalyptus offers our customers a complete cloud platform solution starting from private cloud, spanning into hybrid model and running PaaS on top of it. This is a great opportunity for very large enterprise customers to have their own “Private PaaS.”

OrangeScape provides a model driven visual development environment for creating cloud based business applications that can be deployed on the OrangeScape run-time platform and the cloud infrastructure of their choice, such as Google App Engine, Microsoft Azure, or Eucalyptus’ IaaS platform.

Eucalyptus is the world’s most widely deployed cloud computing software platform for on-premise (private) Infrastructure as a Service clouds. The EucalyptusIaaS platform is fully compatible with the Amazon Web Services (AWS) APIs, enabling customers to more efficiently migrate workloads between their existing data centers and AWS while using the same management tools and skills across both environments.

“Through our partnership with OrangeScape, customers can combine the Eucalyptus IaaS platform with OrangeScape’s Visual Platform as a Service (PaaS) to create a production ready private PaaS for enterprises,” said David Butler, senior vice president of marketing at Eucalyptus. “The integrated solution meets the market’s increasing demand for comprehensive cloud solutions, and will provide customers the fastest way to bring business applications and services to market.”

Through this partnership, OrangeScape will provide technology platforms for customers globally to build, deploy and scale sophisticated business applications, with flexibility to opt for a an Enterprise Private Cloud or Hybrid Cloud configuration.

Step 4 – Exporting the database and deploying WordPress

Now that we have the environment configured in AWS Beanstalk and the MySQL database created in Amazon RDS, it is time for us to move WordPress to the Cloud.

First, we will export and import the MySQL Database. I am using the MySQL Workbench for this. But you can use either command line or PHPMyAdmin to export and import the database.

With the database exported to RDS, we are all set to move WordPress to AWS Beanstalk. But one final step is to change the wp-config.php to point WordPress to the RDS DB Instance. We will replace localhost with the RDS endpoint in the wp-config.php.

Navigate to the WordPress folder and run the following commands

Git add .
Git commit – m “final check-in”

The above command should result in the following indicating that the wp-config.php change is now committed.
1 file changed, 1 insertion(+), 1 deletion(-)

We are almost there! Time to run the last few commands to finish the deployment. Run the following commands to complete the deployment. Carefully enter the details that match with the AWS Elastic Beanstalk configuration that we created earlier.

git aws.config

Finally, take a deep breath and run the following command

git aws.push

This will take a while as it uploads all the files to the AWS Beanstalk environment. After the upload is complete, make sure that the environment is healthy and green.

Clicking on the URL should open the WordPress blog running within AWS Elastic Beanstalk.

Congratulations! You have successfully deployed WordPress on AWS Elastic Beanstalk!

Notice that the URL for the WordPress blog is now available at .elasticbeanstalk.com

PS – If you have issues with rendering the blog correctly in the browser, log into the WordPress dashboard or change the settings in wp_options table to update the site settings to reflect the current URL.

- Janakiram MSV, Chief Editor, CloudStory.in

Step 3 – Creating a MySQL Database in Amazon RDS

Login to AWS Management Console and select Amazon RDS

Select Launch DB Instance

Select the MySQL database and click on select

Choose the default version of MySQL DB Engine Version, db.m1.small for DB Instance Class and No for Multi-AZ Deployment. Choose 5GB as allocated storage and provide an identifier, username and password for the database instance.

Choose wordpress as the database name and click on continue.

Leave the defaults and click on continue.

Review the configuration and click on launch DB instance.

After a while, the DB instance should be available. Make a note of the Endpoint which will be included in the WordPress configuration at
a later point.

Click on the DB Security Groups in the left navigation pane and click on Create DB Security Group

Provide a name and description and click on Yes, Create.

Select EC2 Security Group in the Connection Type and elasticbeanstalk-default in Details and click on Add. This setting allows the requests originating from the AWS Beanstalk applications.

We will also add a security group to allow traffic from any machine. This is to allow us to import the database one time. After this operation, don’t forget to revoke this group.

Click on the Modify button and associate both the security groups with the DB instance.

We created the MySQL database on Amazon RDS for the WordPress deployment. In the next and the last part of this tutorial, we will import the database and deploy WordPress AWS Elastic Beanstalk. Stay tuned!

- Janakiram MSV, Chief Editor, CloudStory.in

Recently Amazon has added the support of PHP and now developers can perform a git push to easily deploy PHP web applications. This brings AWS Elastic Beanstalk at par with other PaaS offerings. This tutorial walks you through the process of deploying WordPress on AWS Elastic Beanstalk. This assumes that you have a setup of WordPress locally on your development machine.

Prerequisites –

• Running setup of WordPress
• Active account with Amazon Web Services
• Git version control system
• MySQL Workbench
• AWS Elastic Beanstalk command line tool

Step 1 – Create a local git repository and initialize the AWS Dev Tools

Navigate to the local WordPress directory and run the following command to initialize the git repository

git init .

Now, run the AWS DevTools from the Elastic Beanstalk command line tool download

AWSDevTools-RepositorySetup.sh

Run the following command to add a new repository to git and commit the changes

git add .
git commit -m "first check-in"

The above command will show how many files are changed and inserted into the repository

938 files changed, 240726 insertions(+)

Now the local environment is almost setup for us to deploy the application to AWS Elastic Beanstalk. In the next step, we will create a new PHP application on AWS Elastic Beanstalk that will act as a placeholder for our WordPress blog.

Step 2 – Creating and configuring Amazon Elastic Beanstalk Environment

Log into AWS Management Console and Select the AWS Elastic Beanstalk tab

Select Upload your own application and click on the Launch Application button

Provide a name to the application and select 64bit Amazon Linux running PHP 5.3. Click on Continue.

Provide an environment name and click on Continue.

Select t1.micro for the instance type and click on continue.

Review the changes and click on Finish. You should see the following message.

After a while, you should see that the application has been successfully configured and is accessible at a URL.

Clicking on the URL should show you the sample application that acts as the placeholder for our WordPress deployment.

In this part of the tutorial, we configured git on our local machine and configured AWS Elastic Beanstalk PHP environment to deploy WordPress. In the next part, we will configure the MySQL database on RDS. Stay Tuned!

- Janakiram MSV, Chief Editor, CloudStory.in