Apache Solr is a Java application that gives you a more full featured search experience than the default Drupal search. It provides tools to allow your users to narrow down the search results and adjust the sorting. Solr can be very confusing initially due to the new terms it introduces as well as the sheer number of configuration options. This guide helps you plan out the items that you need to think about before you ever install the module.
Don't worry if this doesn't make sense, we have an upcoming webinar that should help fill in all the blanks. Be sure to check out our free Intro to Solr in Drupal webinar.
1) Identify who will be hosting your solr instance
Solr is a Java application that must be continuously run on a server. Most budget hosts won't allow you to run extra software like Apache Solr on your LAMP account. You'll ether need to run a dedicated server, a VPS or a company with a hosting plan that includes Apache Solr. If you don't have in-house IT I would recommend you go with a hosting plan that includes Solr.
2) Make sure your Solr version is compatible with the Drupal module
There are three major branches of Apache Solr today. 1.x, 3.x and 4.x. The good news is all three branches are supported for the Apache Solr Drupal module for D7. If possible try to go for the latest version of Apache Solr since they usually include performance enhancements. The Drupal 7 module recently added support for Solr 4.x so make sure you have the latest version of the module.
3) Figure out which content types/entities/bundles you want to have indexed
The Drupal module allows you to choose which entities (also known as content types) you want to have indexed. This can be helpful so you can keep information out of your search index that you don't want users finding. An example of something you don't want in the search results might be Slideshow slides or nodes that only contain a link to a movie file.
4) Figure out which facets you'll have on search
Facets allow your users to choose the information that is relevant to them from the search results and then narrow down the results. You've probably used facets on popular shopping sites without even knowing it (hint: You narrowed down the search for items between $50 - 70 and only available for free shipping.)
In the Drupal 7 version of the module you'll need to download Facet API to have facets. This will allow users to narrow down by Author, Content Type, Language, Post date, Tags, Updated dates and many other data points. Don't forget to add the blocks to the search page to be displayed.
Once you get past the initial shock of dozens of new terms Apache Solr can be an easy and effective solution. Solr scales from the smallest of sites to large clusters of sites with minimal effort while still allowing you the level of customization that your company needs.