Kleistsaal (Pink)

This talk will discuss challenges of distributed searching, scoring and ranking, and merging partial result lists. Common issues in such situations include: global IDF calculation, global faceting, assigning ranks when merging partial lists, etc. Several effective approaches to these problems will be presented using Solr as a search platform.

So you have downloaded Solr, configured it, indexed your data and are now ready to integrate it with the rest of your enterprise Java application. For most situations, this process will begin with Solr's Java client library SolrJ, but how do you configure SolrJ? What are best practices for using it? And how can you use it to build a robust search service tailored to your applications needs? In this presentation we'll show how to configure SolrJ in a Spring application and walkthrough a powerful design for incorporating Solr/Lucene and SolrJ in your application.

The Mozilla Metrics team in their role as data services for Mozilla is developing a new data storage and processing infrastructure for the Socorro crash report system.

HBase has been deployed in production to solve often very different problems at companies such as StumbleUpon, Mozilla, Facebook, and Twitter. It helps scaling web applications, it stores insanely high amounts of historical data, it’s used as source or sink of analytical jobs, it keeps track of millions of counters, and so on. This presentation aims at helping people interested in HBase and current users understanding how it can be adapted to their own use case by analyzing the current deployments and their characteristics.