There is great excitement about the cloud; however, it is important to conduct a full assessment as part of the Cloud Transition.

It is very important to decide which applications will be moving to the cloud based on factors such as alignment with business requirements, cost savings, level of effort, available resources, benefits of the new solution, risks etc to determine business and functional benefits. Architecture and security aspects need to be thoroughly assessed, and factors such as interoperability, reliability, portability, storage, data transfer, sensitivity of data have to be evaluated.

The initial and projected cost implications for services have to be monitored regularly since services that are cost effective initially may end up being more expensive down the road as more users are brought on board. As part of the assessment of cloud vendors, technology and cost aspects need to be looked into carefully and expectations should be documented in the service level agreements.

Many organizations have started with small applications such as backup capability or test, development environments to reduce the migration risks. Email, collaboration software, non-mission critical applications have been good candidates for initial transition to the cloud. Development and test environments can be built and brought down quickly using the cloud, applications that have a heavy load for certain times are also good candidates. For example advertising, marketing, information collection and dissemination capabilities that are only needed to be running during a certain time of the year are such candidates.

The suggested approach is to identify and migrate low risk applications initially to the cloud. There may be mission critical applications that have been in production for many years and these require more thought before moving to Cloud solutions. Some such applications may not be easy to re-architect and interface with new services, especially if they have data exchange mechanisms and interfaces to other applications that need substantial architectural changes. In some such cases, the answer may be - if it ain't broke don't fix it!

The transition to the Cloud may also require managing Big Data effectively. Big data requires adequate storage and distributed processing. Big Data also injects heterogeneous data types that need proper integration into the existing infrastructure. Social networks have been generating tremendous amounts of such data. Big Data is truly a nightmare and it is important to effectively manage such data. Big Data tools can manage and analyze structured and unstructured data and can obtain information from large data sets. Software integration platforms can support Big Data with the Cloud and software frameworks, web services can support data access, updates and storage. To promote reuse the datasets should be linked to strong metadata capabilities. Big data can be located over many servers and hence compiling the data should be managed across these servers. Big data may also require parallel processing and software running on multiple servers.

Generally, a phased approach across the enterprise may be very useful, rather than moving everything at once, so that the risks can be effectively managed.  It may be easier to start developing new capabilities by leveraging the Cloud and they should be built by keeping in mind key considerations for interfacing with Cloud services.

(This has been extracted from and is reference to Ajay Budhraja's blog)

• Cloud Transition Methodology - Approach and Best Practices
• Booz Allen Hamilton Lays Out Path To Cloud
• CLOUD² Commission: Developing Your Cloud Transition Plan