One of the best things about Drupal’s open-source ecosystem is that it empowers you to be open-minded. Given the vast array of solutions and modules available, users can customize their site to their whims. Alternatively, if you think up and code something new, your contributions can be shared online with other users. With all of the customization available, Drupal is a conducive platform for outside-the-box thinking.

Decoupling is a recent example of this philosophy. Where a standard Drupal website would feature a Drupal-powered front and backend, decoupling opens the door for a variety of possibilities. A decoupled site can utilize different platforms and technologies for both the front and backend. For example, a decoupled site could utilize Drupal’s backend CMS while running a React-powered frontend. Such is Drupal’s flexibility that it can power scores of different, user-facing channels from a single backend, including other sites, native apps, Internet of Things (IoT), and more.

This decoupled or “headless” concept has more applications than just for site design, though. The search function of a website, for one, can benefit from components that utilize this headless approach – and not a moment too soon. As Google has begun to sunset its Google Search Appliance offering, there is now a need for an open and flexible search tool with enterprise-level capabilities.

At this year’s Midwest Drupal Camp, the team from Palantir demonstrated that a decoupled approach to site search was viable. This solution, federated search, allows for indexing and searching across multiple sites. For organizations with a large web portfolio across different platforms, this open federated search solution can fill the gap left by Google.

Understanding why federated search for Drupal is important requires an understanding of how regular site search functions operate. At the core, the search feature is built from three different components: the source, index and results. The source simply refers to all of the searchable content on a given site, from blogs to landing pages. The index is a compilation of metadata that makes the content form the source easier to parse. At Duo, we often use Apache Solr, a platform-agnostic, open source solution for indexing, as it provides speed, power and its own server capabilities. Finally, the results refers to the front-end experience that compiles and delivers the search results to the user.

The above setup will work fine for most simple websites, but larger organizations often require a more robust solution. With federated search, users can query across multiple sites across different platforms without placing much strain on Drupal, since Apache Solr is handling generating the index and providing results. This is accomplished through some tweaking of the basic site search formula.

Part of what makes this search so powerful is that it takes advantage of Drupal’s backend without relying on its frontend. For that, Apache Solr’s dedicated servers empower this new search solution by shouldering the burden of indexing and providing the results. Before it can work, though, some configuration is needed. Based on this configuration, Apache Solr can encompass searches across different sites – including sites that aren’t built with Drupal. Creating this custom solution, in conjunction with the Search API and Search API Solr modules, will ensure that the different data types being indexed will be standardized.  

As for the results section, the best approach is a decoupled one. Building out the front-end component in the React JavaScript library allows for robust searches without slowing down the rest of the site. By using Drupal’s CMS, Apache Solr and React’s power in coordination, any organization can create a search feature that quickly indexes vast ranges of data and delivers it in an easily digestible manner. For a deeper dive, Palantir has made their demo of federated search available.

This powerful and streamlined take on site search has a variety of applications. Before releasing the solution, Palantir originally developed federated search for the University of Michigan, as each department ran their own sites on different platforms. Federated search now allows users to seamlessly search for information across the entire school’s network, regardless of the technology used to deliver the content. Beyond university ecosystems, federated search also presents an opportunity for eCommerce. Using this solution, products from different vendors can be consolidated into a simple search.

Thanks to Drupal being open source, organizations can utilize federated search and any other contributed solution at any time. This level of openness is what makes Duo such champions of the Drupal platform. At Duo, we’re committed to exploring new features like this and helping each of our partners think outside the box. If you’re ready to start rethinking your website or sites, we’re just a click away.