New Whitepapers on Cloud Migration: Migrating Your Existing Applications to the AWS Cloud

Developers and architects looking to build new applications in the cloud can simply design the components, processes and workflow for their solution, employ the APIs of the cloud of their choice, and leverage the latest cloud-based best practices for design, development, testing and deployment. In choosing to deploy their solutions in a cloud-based infrastructure like Amazon Web Services (AWS), they can take immediate advantage of instant scalability and elasticity, isolated processes, reduced operational effort, on-demand provisioning and automation.

At the same time, many businesses are looking for better ways to migrate their existing applications to a cloud-based infrastructure so that they, too, can enjoy the same advantages seen with greenfield application development.

One of the key differentiators of AWS’ infrastructure services is its flexibility. It gives businesses the freedom of choice to choose the programming models, languages, operating systems and databases they are already using or familiar with. As a result, many organizations are moving existing applications to the cloud today.

In that regards, I am very excited to release our series of whitepapers on cloud migration.

Download Main Paper : Migrating your existing applications to the AWS cloud (PDF)

This whitepaper will help you build a migration strategy for your company. It discusses steps, techniques and methodologies for moving your existing enterprise applications to the AWS cloud. There are several strategies for migrating applications to new environments. In this paper, we share several such strategies that has helped enterprise companies to take advantage of the cloud.  We discuss a phase-driven step-by-step strategy for migrating applications to the cloud (see below).

 
To illustrate the step-by-step strategy, we provide three distinctly different scenarios listed in the table. Each scenario discusses the motivation for the migration, describes the before and after application architecture, details the migration process, and summarizes the technical benefits of migration: 

Scenario Name	Solution	Use case	Motivation For migration	Additional Benefits	Services Used
Company A	Web Application	Marketing and collaboration Web site	Scalability + Elasticity	Auto Scaling, pro-active event based scaling	EC2, S3, EBS, SimpleDB, AS, ELB, CW, RDS
Company B	Batch processing pipeline	Digital Asset Management Solution	Faster time to market	Automation and improved development productivity	EC2, EBS, S3, SQS
Company C	Backend processing workflow	Claims Processing System	Lower TCO, Redundancy	Business continuity and Overflow-protection	EC2, S3, EBS, AS, SQS, IE

1. Migration Scenario #1: Migrating web applications to the AWS cloud (PDF)
2. Migration Scenario #2: Migrating batch processing applications to the AWS cloud (PDF)
3. Migration Scenario #3: Migrating backend processing pipelines to the AWS cloud (PDF)

As always we are hungry for customer feedback. Let us know whether this paper is helpful to you. If you have moved an existing application to the AWS cloud, we would love to get your feedback too. Send us more details at evangelists at amazon dot com.

- Jinesh Varia

P.s. More AWS whitepapers are available at http://aws.amazon.com/whitepapers  

 
 
 



Comments

[Miha Ahronovitz]

"We should make the things as simple as possible, but not simpler", is a quote from Albert Einstein. This white paper covers all aspects of the migration, and this has intrinsic complexity. The Business development people want to know more data before being able to sell a migration to AWS idea. I will comment about this. You exposed additional business model, besides the pay per use: BYOL (bring your own license) and ISV-cloud based SaaS. This is the most interesting thing about migrating to AWS: billing. Not only one receives a bill for the usage from AWS, but the customers can use Amazon billing capacity to send invoices to users of applications provided via AWS. The focus on the economics of migration, much more than the reduction in TCO, is the ability to maximize revenues. Once migration is completed, the organization can bill everyone, customers and internal users. Even if the internal users do not pay, one has a clear record and how it costs to provide internally the services.

So the idea to reduce TCO is out of context. We do not want to reduce TCO. If it were, one better shuts down the data center & the cost will be zero. This is , of course, absurd. One want to maximize the revenues, as expressed by ROI. If I have the bill I pay for AWS, and the bill my customers pay me, I have an instant ROI. I can improve it by adjusting usage up and down, and by variation up and down in the prices charges for applications delivered as service. Viewed in this light,, the migration to AWS transforms the customer data center from a cost center into a profit center. This should be the biggest incentive to migrate, more than anything else. See my slides http://bit.ly/cfj6sz

A second consideration are is the security. Actually the security of public clouds, statistically speaking, is superior to the security of 80% of the private clouds up there (Gartner). See also my post on Google Cloud Group http://bit.ly/fYpLpX . A recent example is #wikileaks, which moved quickly to AWS ro escape denial of service attacks.

Immediately after this WP was published, Adrian Cockcroft from Netflix presented "Netflix on AWS" migration, a key reference on how a 16 million subscriber company can operate almost 100% from AWS. http://blip.tv/file/4252897

Jinesh, your white paper is a tremendous effort in the right direction. I mentioned before the clarity and elegance of each WP you write. You strike oil, now we need to go and take it out of the ground

[Jinesh]

Very insightful comment, Miha and Thanks for the nice words. This is a team effort and hence I cannot take all the credit.

I agree with you that usage-based costing model of the cloud is certainly one of the motivations of migration. I will just add a small excerpt from the best practices paper Usage-based costing: With utility-style pricing, you are billed only for the infrastructure that has been used. You are not paying for allocated but unused infrastructure. This adds a new dimension to cost savings. You can see immediate cost savings (sometimes as early as your next month's bill) when you deploy an optimization patch to update your cloud application. For example, if a caching layer can reduce your data requests by 70 percent, the savings begin to accrue immediately and you see the reward right in the next bill. Moreover, if you are building platforms on the top of the cloud, you can pass on the same flexible, variable usage-based cost structure to your own customers. The blog post that I wrote a year ago discuss a few things related to that front http://aws.typepad.com/aws/2009/07/the-cloud-as-a-platform-for-platforms.html. Outsourcing companies can take advantage of the cloud and actually only bill the clients for the infrastructure they actually consume for the project.

While "reducing TCO" (and maximizing revenue) is an important driver, we are seeing that there are different motivations for migration. Netflix move was primarily to reduce time to market. NASA move to AWS cloud is because they can run more missions and do more research and reduce their backlog of innovative projects (read "launch new missions quickly"). Hence I typically advice customers to define their success criteria before hand.

Thank you for commenting and tweeting about the paper.