Founder & Project Lead at JBoss Cache, Infinispan Data Grid
By Adrian Cole
I met Manik in New York a little over two years ago. He eloquently taught advanced clustering technology to a select few who flew in for the opportunity. Since then, I've recognised that Manik is a true renaissance man. From Mobot to JBoss, he leads in both designing and evangelizing technology, and is capable in countless contexts. Manik co-founded jclouds while at the same time launching a cloud-friendly grid solution Infinispan. If there is someone to watch out for in the cloud industry, Manik Surtani will keep your interest, and keep you current.
Never before has data storage and retrieval been more at the forefront of what's new and exciting in technology, and it's long overdue for a major update. Consider, how much have filesystems or rational databases really changed over the last two decades? Web 2.0 has introduced a whole class of massively distributed, socially spread applications which impose unprecedented volumes on data storage systems.
In this talk, Infinispan founder and Jboss Cache project lead Manik Surtani introduces the role of data grids in today's cloud computing environment. The extreme scalability offered by data grids are powering the greatest and most high profile of today's applications, and data grids take on a far more prominent role in cloud deployments as traditional databases hit scalability and resilience issues. In this talk Surtani introduces Infinispan, the new open source data grid platform, and the motivations and evolution as a project. Distributed data structures and the use of data grids as a viable, cloud-ready data storage mechanism is discussed in depth.
Manik talks about how Infinispan is heralding in a new era in Open Source Data Grids and how it compares to Oracle's Coherence. He also talks about the future of Infinispan and how developers can get involved.
Manik Surtani explains the importance of parallelism and covers the different types of cache and clusters. Moore's Law, from Intel co-founder Gordon Moore, ensured that the rate at which computers got faster was enough to provide performance gains that can automatically be realized by an application to cope with growth in demand. Moore's law is no longer relevent, today, Armdahl's Law, coined by Gene Amdahl, is more applicable. This law says that by throwing more processors to work on a problem, you don't get the expected overall throughput calculated by adding the individual throughputs of each processor. A problem can be broken down into subproblems and then distributed to multiple processors and the effort in collecting results will have an effect on the overall throughput.