This site is archived.
Abstract: 

The search revolution is finally available to anyone with a Drupal site. By connecting Drupal to a Solr search backend, new possibilities arise for how to structure sites, discover content, and keep your visitors coming back. Our case study will be Drupal.org and Groups.drupal.org. It's the community you know and love, but you're going to love it even more when you can search and drill down into groups, authors, tags and dates, then save your search and pipe it into an RSS feed.

Presenters: 

Jacob Singh
Robert Douglass
Peter Wolanin
Audrey Foo

More Information

The search tool can become the central interface though which visitors find information on your site - supplanting the need for extensive hand-crafted site sections and navigation menus and improving the user experience by letting them quickly find what they are looking for following a pattern that is natural to them. Unfortunately, a SQL database is not optimized to provide these sorts of features, so solutions built totally within the PHP/SQL framework have difficulty performing well for large sites or for multi-sites search. Indexing the data from site content with Apache Solr allows you to offload the computational effort and optimization to a separate server. We will show how Drupal integrates with Solr and how a search integrating content from both groups.drupal.or and drupal.org could look.

Agenda

  • Brief introduction to the project and presentation.
  • Powerpoint presentation
    • Why search is broken on d.o. and g.d.o (everyone uses google, shh...)
    • It's all about facets.  Text is only the first step.
    • Recommended content using Solr
    • Multi-site search.  What it works for, what are the limitations
    • Bonus features:
      • Spell Checking
      • Highlighting
      • Phonetic search
  • Live Demo and trials from the audience
    • Get suggestions of Drupal research people need to do in the audience, we'll run the same queries on core search, and our search.
  • Presentation on setup / installation / infrastructure
    • Overview of Solr – How data is indexed, how it is retrieved
    • Brief look at the admin interface for Solr
    • Installing Solr / Java
    • Installing the schema / AS module
    • Indexing Data
    • Configuring facet blocks

Goals

  • Attendees will find out how Solr can improve the usability of their site

  • Get a solid overview of the technology involved without going into the bloody details

  • Developers will find out about BoFs happening around Solr and will attend

  • Provide an brief introduction to Acquia's hosted search service

### Resources

Nothing, just a painful experience with search (easy to come by) and a tolerance for bad puns.

### Speakers (not yet finalized)

Robert Douglass

Robert’s Drupal adventure started in 2003. Along the way he co-authored the first book published about Drupal, was technical editor of Pro Drupal Development, became the maintainer of numerous modules, and organized Drupal’s “Google Summer of Code” involvement. Robert loves classical music and open source software dearly, and looks to each as a source for motivation and optimism.

Jacob Singh

Jacob joins Acquia after many years of consulting as a system architect and project manager specializing in open source software for non-profits, enterprise software planning and agile development. He started with Drupal in 2006 with CivicActions where he led the effort to build Amnesty International's Drupal site. His work these days is focused around building scalable applications with Amazon Web Services and excellent search products. Originally from Western Massachusetts, he now lives in New Delhi with his wife and tries to get in Celtics games on the internet in the middle of the night, pickup basketball and a bit of fooling around on the guitar while not, against better judgment, tinkering with some bleeding edge FOSS product.

Peter Wolanin

Peter Wolanin made the leap (or perhaps just one step in a biased random walk) from being Senior Scientist at a bio-tech to working on what he truly loves, Drupal. Peter spent a large fraction of his non-working waking hours in the last couple years re-working Drupal core and maintaining contributed modules. In the free time that he hasn’t devoted to Drupal he enjoys watching and reading science fiction, hiking, getting involved in politics, and working on his town’s environmental commission. Peter graduated Cum Laude from Princeton University with an A.B. in Physics and a certificate in Engineering Physics and then went on to obtain his Ph.D. in Physics at the University of Michigan where his research focus was biophysics.

9 Comments

The slides are great - and would provide great talking points for my (reluctant) managers. Can you provide a link to them?

Since this talk was rolled into a panel discussion, we will also have a more focused BoF on Solr: http://dc2009.drupalcon.org/node/3836

whereas the presentation will start with an overview for people unfamiliar to the technologies. We hope you all come to both =)

Will ApacheSolr run on my own server ? Will extra softwares make things more difficult on already available cpu resources ?

Many users are looking for Forum solution to vbb or phpbb integrations, again for search there is yet another ApacheSolr. So my question is core drupal only about user managment and custom content types ?

Looking forward to this one.

+1

Don't think we'll be covering the exact same info, but perhaps we can exchange powerpoints a couple weeks before the con?

http://dc2009.drupalcon.org/session/google-custom-search-engine-solr-roa...

This sounds great! I can't wait.