LJC - London Java Community - Scaling Big Data Pipelines At The BBC/A Practical Approach To Java Memory Model
Thursday 5 April 2018, 19:30 - 20:30
Sign up on Eventbrite for Location, London, Reino Unido
This is a placeholder for the event which is being run on Eventbrite. See the full details & register here: https://www.eventbrite.co.uk/e/scaling-big-data-pipelines-at-the-bbc-by-moving-to-serverless-a-practical-approach-to-java-memory-tickets-44255209523 In December the LJC & Devoxx UK ran a clinic to help people in the community who are interested in speaking at events. Many members of the LJC, including the likes of Trisha Gee & Daniel Bryant, started their speaking career at an LJC event, & can now be found at events around the world. This is something we are committed to continuing. Two attendees of the December event were encouraged enough to submit a proposal to Devoxx UK, & we’re pleased to say they both made it into the agenda. Join us as they debut the talks they will be giving at Devoxx UK for the LJC. Title: Scaling big data pipelines at the BBC by moving to serverless Summary: The BBC Audience Platform Data team consumes, transforms & delivers data from a variety of sources, including BBC Account, the BBC User Activity Service & from audience interactions with the BBC web site and iPlayer. We will take you through the journey we have been on migrating from clusters & Map-Reduce towards serverless compute with the goal of real-time analysis of the billions of events that occur daily. We’ll cover what has worked well, and perhaps more importantly, what has not. Speakers: Blanca Garcia-GilI am a senior software engineer at BBC. I currently work with big data pipelines & I am very interested in automation (testing and deployment) & monitoring of our systems. Neil CroftsNeil is a Software Engineer in the BBC Audience Platform Data team responsible for processing & analysing the billions of messages audiences create when interacting with BBC online services. He also worked extensively with XML based technologies on the BBC Red Button Digital Text service. Previously he spent his time developing software in the Automotive, Telecoms & Electronics industries for a number of blue chip organisations. Title: A practical approach to Java memory model Summary: Java Memory Model is an interesting area of multi-thread programming. It describes how threads interact through memory & what behaviours are allowed in concurrent execution. Most java developers are familiar with mutual exclusion & synchronized keyword but sometimes we forget that beside mutual exclusion we also have to care about memory visibility. The talk will introduce number of examples to help understand JMM. It will use OpenJDK Java Concurrency Stress tests framework (JCStress) to validate the results. Speaker: Andrzej CzarnyAndrzej is a Senior Software Engineer at Ocado Technology. His work is mostly focused on real time performance of warehouse control systems. He is involved in creating testing frameworks & Continuous Integrations for multiple teams. His main area of expertise is Java & concurrency. This is a placeholder for the event which is being run on Eventbrite. See the full details & register here: https://www.eventbrite.co.uk/e/scaling-big-data-pipelines-at-the-bbc-by-moving-to-serverless-a-practical-approach-to-java-memory-tickets-44255209523
Publicado por: Betabeers