For general requests regarding the EuroSys 2011 tutorials, please contact the tutorial chair Adam Dunkels (contact: firstname.lastname@example.org).
EuroSys 2011 will host the following tutorials on April 10th, 2011:
|T1||Data Deduplication - Algorithms and Implementations|
|T2||Distributed coordination using ZooKeeper|
Data Deduplication - Algorithms and Implementations
Andre Brinkmann (University of Paderborn)
This tutorial presents and discusses the different approaches laying behind today’s data deduplication environments. Data deduplication systems discover redundancies between different data blocks and remove these redundancies to reduce storage capacity. Data deduplication has become a hot topic in both academia and industry, as it promises a huge reduction of the required backend storage capacity.
Distributed coordination using ZooKeeper
Flavio Junqueira (Yahoo! Research), Benjamin Reed (Yahoo! Research)
In this tutorial, we ﬁrst review the basic challenges of distributed systems from both a theoretical and real world perspective to motivate the need for a coordination system. We introduce ZooKeeper and show with examples how one can use it to simplify the design of a distributed system. Even though ZooKeeper enables the system developer to deal with challenges of distributed systems eﬀectively and eﬃciently, it does not eliminate them completely. We also review some of the corner cases that designers and developers still need to deal with. We will also share some of the pitfalls that others have fallen into, and explain why some obvious new features, such as watches with information about changes, are bad ideas. We will conclude by sharing upcoming changes, desired features, and ZooKeeper related research directions.