Skip to Content

You are currently visiting an old archive website with limited functionality. If you are looking für the current Berlin Buzzwords Website, please visit https://berlinbuzzwords.de

Urania Berlin, June 6-7, 2011

Einsteinsaal (Green)

Node.js for heavy I/O

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 11:00 - 11:40
Speaker: 
Felix Geisendörfer

"Server-side JavaScript has been around since 1996, but despite the huge success on the client, it never quite managed to attract a significant audience.

This all changed in 2009, when Ryan Dahl presented his non-blocking I/O framework for Google's V8 JavaScript engine at JSConf.eu. The radical new approach, combined with one of the fastest JavaScript engines, has attracted an explosive number of early adopters ever since.

This talk gives you a brief introduction to node.js, exploring the technology behind the buzz.

Advanced Routing Patterns in ZeroMQ

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 11:50 - 12:10
Speaker: 
Daniel Trümper

This talk will dig into the messaging patterns of ZeroMQ, a "socket
library, that acts as concurrency framework". We will start by using ZeroMQ messages for creating a Least Recently used router. Using a ZeroMQ device we will then create a N-M router that serves as proxy between many masters and many slaves. Finally the Mongrel2 messaging architecture will show how to emulate an asynchronous request/response pattern.

Building next-generation Web Apps with WebSocket

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 12:10 - 12:30
Speaker: 
Matthias Wessendorf

To enable Sever-Side-Push in Web-Applications a lot of hacks (aka Comet/Bayeux) have been done in the past. The WebSocket spec fixes that by introducing a bi-directional and full duplex communication channel over a single TCP connection. This session gives an overview of the WebSocket API and shows how to use other protocols like AMQP or Stomp on-top of it, to build powerful and future-proof Web-Applications.

Messaging Patterns with RabbitMQ

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 13:30 - 14:10
Speaker: 
Alvaro Videla

RabbitMQ is emerging as good solution for open source messaging in modern architectures. This opens new challenges on how to implement integration patterns in order to go beyond simple produce/consume applications. In this presentation we will show how to implement several messaging patterns using RabbitMQ as backend technology such as:

- Competing consumers
- RPC
- Parallel RPC
- Smart Proxy
- Publish/Subscribe

The examples will be give using standard AMQP commands so the attendee can translate them later to her favorite RabbitMQ/AMQP client.

The other Apache technologies your big data solution needs!

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 14:20 - 15:00
Speaker: 
Nick Burch

You've gone to the talks on Hadoop / SOLR / NoSQL / etc, and now you're ready to start building your own solution on top of that! What you might not realise is that you may end up reinventing some bits of the wheel whilst building your system..

In this talk we'll take a whistle-stop tour through a number of the projects from the Apache Software Foundation that aren't "big data" projects, but which could prove very helpful to you in building your big data solution.

Time Series or Causal Analysis Without Limits!

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 15:20 - 16:00
Speaker: 
Shevek

Analysis of a causal or time series relationship between two data sets (or functions) is important for fields from yield optimization to signal processing, stock market analysis to functional genomics, and many other applications.

This talk describes an algorithm developed by Karmasphere Labs for performing the entire family of cross correlation algorithms on arbitrarily large data sets. The algorithm supports wide or even unbounded windowing functions.

Real-time Analytics With Cassandra

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 16:10 - 16:50
Speaker: 
Sylvain Lebresne

Real-time analytics requires counting a lot of things very quickly.
This talk will present the Apache Cassandra approach to distributed
counters, starting with the difficulties inherent to partitioning
counters across multiple machines, and continuing to the solution
adopted by Cassandra. We will wrap up with a use-case: how Twitter
leverages this counter design in their Rainbird system for real-time
engagement analytics.

ZooKeeper - the unsung hero

Location: 
Einsteinsaal
Date and time: 
Mon, 2011-06-06 17:00 - 17:40
Speaker: 
Thomas Koch

ZooKeeper started as an internal Yahoo project. After being released as free software to the Apache foundation it became a critical building block of well known systems like HBase, Cassandra, Neo4j and SOLR and many more lesser known but critical systems at linkedin, yahoo, facebook and rackspace. Despite being so useful and important, most people are not aware of the hard working ZooKeeper.

ZooKeeper helps you to supervise the availability of processes (heartbeat), to implement distributed locking, leader election, central configuration and reliable messaging.

Heavy Committing: Flexible Indexing in Lucene 4.0

Location: 
Einsteinsaal
Date and time: 
Tue, 2011-06-07 11:00 - 11:40
Speaker: 
Uwe Schindler

Apache Lucene's next major release, 4.0, will introduce lots of flexibility into indexing, but also fundamental changes to the well-known APIs: It features a new and consistent, 4-dimensional iteration API on top of a low-level, pluggable codec API giving applications full control over the postings data. Terms are now arbitrary opaque bytes enabling users to store terms in any encoding, not necessarily UTF-8, natively in the index (e.g. numeric fields).

Heavy Committing: Lucene DocValues aka. Column Stride Fields

Location: 
Einsteinsaal
Date and time: 
Tue, 2011-06-07 11:50 - 12:30

This talk will introduce one of the new features in the next intentionally backwards incompatible Apache Lucene release. Beside Flexible Indexing & Realtime Search Lucene 4.0 will ship with an alternative column based per-document storage called DocValues or Column Stride Fields. DocValues provide efficient storage for per document values used for sorting, result rendering or scoring. Unlike FieldCache DocValues don't need to be indexed nor be un-inverted and allow users to specify how values are stored per field.

Syndicate content