Speaker: Manik Surtani, JBoss

Presentation topic: Distributed Caching – Preparing for an Era of Highly Available Concurrent Computing

Language: English

Manik Surtani is a core R&D engineer at JBoss, a division of Red Hat, and project lead on JBoss Cache. He has a background in artificial intelligence and neural networks, a field he left behind when he moved from academic circles to the commercial world. Since then, he’s been working with Java-related technologies, first for a startup focusing on knowledge management and information exchange, and later for a large London-based consultancy as a tech lead focused on e-commerce applications on large Java EE and peer-to-peer technology. Surtani is a strong proponent of open source development methodologies, ethos, and collaborative processes.

Abstrakt: Grid computing is becoming increasingly popular, particularly with  cheap commodity hardware, and frameworks that make building otherwise-complex systems easier.  The change in the way CPU manufacturers scale  CPUs - resulting in a shift where Amdahl’s Law gains importance and  Moore’s Law no longer applies - have also meant that people need to change the way they write software. Concurrency becomes ever more  important and queueing and synchronization - which have always been  pain points - are further exacerbated.  As a result caching becomes  more important as touching the disk becomes an increasing bottleneck.

This is an in-depth discussion about achieving performant and  effective distributed caching, whether for failover and high  availability, removing bottlenecks such as databases, or enabling the distribution of state for compute grids.  Among topics covered are  minimizing latency across a network and optimizing how data is located and processed. The discussion will also go into the tradeoffs between maintaining coherence and resilience.