I am at the 7th ACM International Conference on Distributed Event-Based Systems in Arlington, Texas.
On Sunday morning I attended the tutorial entitled Stream Processing Optimizations, by Scott Schneider and Martin Hirzel, IBM Thomas J. Watson Research Center and Buğra Gedik, Computer Engineering Department, Bilken University. They covered terminology of stream processing, a catalogue of possible optimisations, and then an in-depth analysis of how to do the “fission” optimisation with applications written in IBM’s SPL.
The afternoon began with Personal Big Data Management in Cyber-Physical Systems – the Role of Event Processing by Nenad Stojanovic, Ljiljana Stojanovic, and Roland Sthmer, all of FZI Forschungszentrum Informatik. The content was about building event-based systems that handle data about individuals; I didn’t find it very illuminating or interesting.
The day ended with me going to part 2 of Event-based Systems Meet Software-Defined Networking by Frank Drr and Boris Koldehofe, Institute of Parallel and Distributed Systems (IPVS), University of Stuttgart. (Well, that’s not quite true, as a Texas-sized chicken fried bacon cheeseburger (!) followed.) This was good, talking about how to populate flow tables for OpenFlow with data that effects content-based publish/subscribe. There are interesting optimisations here, partly tied up in how one constructs bit strings for the flow table that are derived from subscriptions and advertisements. Neat stuff, and has given me ideas for a couple of papers.
That’s it for Sunday!