Attended Cloud Camp 2010

I attended Cloud Camp 2010 in Bangalore. I found a good learning experience and this was the first time I attended unconference and Dave Nielsen made the attendance as worthwhile experience. At an unconference, none of the sessions are chosen in advance and the attendees would rise to the challenge and propose topics for discussion.

Hadoop is software platform(Apache Open Software project) that easily lets one easily store and run applications to vast amount of data. The platform helps you to build your own private cloud. Yahoo is making use of this platform to enable providing a private cloud for  their Production and Research teams to innovate. Yahoo  provide source code used in their private cloud to the open source project by sharing stable builds and unearthing bottlenecks and bugs in Hadoop and is available at Yahoo! Distribution of Hadoop. If you are a small enterprise and are looking for a cloud platform to deploy your applications, I would surely recommend to make use of Hadoop and evaluate how your application would scale in a cluster of 100 modes with minimal cost and effort. By the way, Yahoo does not provide a public cloud offering based on cloud. Hadoop exposes all the three layers services,storage and processing are integrated and are exposed by Hadoop. if you want to start development on Hadoop , there is a book called Hadoop: The Definitive Guide

Services<—>Storage<——-> Processing(Scheduling,Mining, Modelling and Data Warehousing).<——>Services

Some  perspectives from Cloud Camp 2010 participants.

  • CloudEra is public cloud enterprise data platform built on Apache Hadoop
  • Cloud has the potential to become the back end processor for the convergent platform of PC ,Television and Mobile
  • Cloud Storage might be different from RDBMS? RDBMS is good only when there are large number of transactional read and writes to a storage and there is a need for random access of dynamic data. if you application has more static data and very little dynamic data, RDMS might not always be the best fit .
  • Grid is platform build to optimize usage of available free processors available in network of processing node. Cloud is way to get resources when you need and is self serviced and you procure procure the needed processor on demand.
  • Power Pivot for Excel 2010 is a data analysis tool that delivers unmatched computational power directly within the application users already know and love—Microsoft Excel and can be used for data mining and data warehousing needs on large amount of data. I would check this when i move to Office 2010.I saw the features and they seem to reflect a superset of Pivot table in Excel
  • I  enjoyed “Introduction to Cloud Computing” by Dave Nielsen and he explained how Amazon, a book seller became  first largest public cloud provider and  why large enterprises did not move to provide cloud offering and why data centre and web site hosting companies were not the first ones to become a cloud provider.  He also briefed upon the traditional it development structure consisting of business users who took charge of the costs and investments, development team for building the application and system administrator to provide right and optimal IT infrastructure. I was wondering  how this model would accept the cloud era of computing. Now one person can take charge of all these things, which means both fast and high risks.
  • I shared with MS team that there is no simple way for developers to move across to windows Azure and there is little or no help with respect to what type of Azure offering provides benefits for developers and the challenge developers face in  the cloud world. MS team also seems to be of the view that VS 2010 cannot be solely depended for building cloud applications without an actual Windows Azure account.