Digest of the upcoming talks

Let us present you with a digest of the upcoming talks that were announced this week.

Easy Statistical Analysis in PostgreSQL with PL/R

PL/R is a PostgreSQL extension that allows the use of R from within PostgreSQL for advanced analytics in a simple, efficient, and controlled manner. It has been available and actively maintained since January 2003. PL/R works with all supported versions of PostgreSQL, and all recent versions of R. It is used by thousands of people worldwide.

Recent trends in big data favor bringing the analytics closer to the data – meanwhile PL/R has been quietly providing this service for over 12 years! This presentation will introduce the audience to PL/R, discuss the pros and cons of this approach to data analysis, take them through basic usage, and finally illustrate its power with a few advanced examples.

Logical decoding: the door to a new world of data exchange and integration applications for PostgreSQL

Logical decoding is one of the most prominent features of PostgreSQL 9.4 but, probably, not the most praised or known. It is the basis for in-core logical replication and many other replication solutions that will be coming to future PostgreSQL versions. But logical decoding is much more than that, it opens the doors to a whole new slew of applications that are not related, or do not require, replication.

Following PostgreSQL's extensibility tradition, logical decoding is pluggable (you can plug in code to determine what the output format is) and, most importantly, it is externally controllable. With this toolset, database DMLs can be streamed to third-party applicationss, enabling middleware solutions for data replication, filtering, modification, routing or interchange with different systems. A new world of data interchange and integration.

Join this talk to learn how logical decoding works, what features does it have, and how to use it from external applications, diving deep into the techniques and the code. Real world examples would be provided and a discussion about new PostgreSQL-based middleware solutions would be started.

PostgreSQL in Brazilian public institutions

In the beginning of the decade of 2000, Brazil has made the choice of using free and open source software whenever possible in the public institutions. One of the challenges was to substitute the well known proprietary choices by PostgreSQL. Every system is big and have high concurrency, a new database system poses several challenges in the data centers, like: how to integrate it with existing hardware, how to have support for it, how to make people believe it will work well and support the load. The formation of technicians and DBAs. During the process, what problems a consultant have to address, give good answers to and how to address them well.

Asynchronous replication uncensored: architectural issues of MySQL, or why PostgreSQL is going to conquer the World

MySQL is a popular RDBMS which is used in numerous projects. As a developer of Percona Server and engineer at Mail.Ru Target, I will tell you about unfortunate solutions for replication in MySQL, explain its internals, showcase the design faults and implementation of multi-threaded replication in the 5.7 version of the server. After this talk you will understand why such replication is a failure, how it should've been implemented instead and why PostgreSQL has avoided these pitfalls.

A JSON representation of DDL

In this session Álvaro Herrera describes my project to allow deparsing of DDL commands to an ad-hoc JSON format, which enables interesting capabilities such as DDL replication, schema auditing, business rules enforcement, and more.