Nuno Preguiça
Associate Professor at DI FCT UNL
Nuno Preguiça is Associate Professor at DI FCT UNL, and leads the Computer Systems group of the NOVA Laboratory for Computer Science and Informatics. The broad aim of his research is to develop techniques that allow applications to efficiently access and share data at any location and at any time. He has been the principal investigator of national projects FEW, Byzantium, RepComp, SwiftComp, the site leader for EU project SyncFree and participated in several other projects. He has received a Google Research Award in 2009 for his work on solutions for cloud data management.
Past Activities
Code Mesh LDN 2018
13.30 - 17.00
Just the right kind of Consistency!
13:30 - 17:00
You need a data store that allows for high throughput and availability, but you are worried about the consistency of your data under concurrent updates when replicating across data centers? Current designs for data storage forces application developers to decide early in the design cycle, and once and for all, what type of consistency the database should provide.
At one extreme, data stores with strong consistency (such as Spanner and CockroachDB) require frequent global coordination; restricting concurrency in this way greatly simplifies application development, but it reduces availability and increases latency.
At the opposite extreme, there are systems like Cassandra that provide eventual consistency only: they never sacrifice availability, but application developers must write code to deal with all sorts of concurrency anomalies in order to prevent violation of application invariants. But the system just needs to be consistent enough for the application to be correct!
In the training day, we will review different approaches and tools for choosing just the right kind of consistency for your app.
OBJECTIVES
You will learn how conflict-free replicated datatypes (CRDTs), transactions, and causal consistency interact to keep your data safe. Further, you will see how to analyse your application and its invariants to obtain a provable correct model. This tutorial is hands-on with many interactive elements to explore the features and limits of the Just-Right Consistency approach.
AUDIENCE
Application developers and system developers.