What s New with Search in Alfresco 5 Mike Farman Alfresco Product Manager Andy Hind Alfresco Senior Engineer
Agenda Server-Side Changes Solr 4 Background Solr Schema Changes Performance New Capabilities Solr 4 Administration UI Share Search Enhancements Live Search Filtered (Faceted) Search Auto-Suggest & Spelling Search Results Actions
Server Side Changes
Solr 1.4 to Solr 4.x A bit of history: Solr releases 1.4 => 3.1 => 4.0 => 4.1.. 4.10 Solr and Lucene Projects merge March 2010 Solr version 4.0: October 2012 Alfresco & Solr: Jan 2012 Major differences from 1.4 to 4.10 New APIs New Schema Indexes not compatible => Mandatory Reindex
Solr 4 Capabilities Advanced Search Capabilities Faceted Search & Filtering Auto-Suggest Spelling Checking Statistics Geospacial Supported with Alfresco 5.0 Date Math Relevance Term-Highlighting Solr Cloud * (Sharding/Replication) Caching (queries, filters, documents) * Old Style Replication is supported
Alfresco 5.0 - Solr 4 Schema Changes DocValues for Faceting and Sorting Lower Memory, Improved Performance Use of Primitive Types Smaller index overhead Spell check and Suggestion support So we can make use of them Improved Cross- Language Support Better precision and recall Site Shortname No path required Query and Facet Better OOTB support for localized analysis and it s customization e.g. Can be changed via schema Does not require new analyzer Geo-properties cm:location aspect stored as geo field (not exposed in UI) Native tag rollups
New Solr Document Store Solr documents are stored on disk All fields and text representation of content Metadata first, then content Binary & Compressed Avoids re-fetching from Alfresco for content index or meta-data only updates Can rebuild index from disk without refetching unchanged nodes from Alfresco FUTURE USE: Term Highlighting
Alfresco 5.0 Model Changes New <facetable> model entry Not required but recommend review custom models before upgrade Fall back rules where <facetable> not specified. <property name= sample:sample_property"> <title>sample Property</title> <type>d:text</type> <index enabled="true"> <facetable>true</facetable> </index> </property>
A note on Solr Cloud Implementation is designed to be Solr Cloud friendly, however, more work is required to support it therefore Solr Cloud not supported for 5.0 However, Old Style index replication is available
New Capabilities Date Math In the last day/week/month/year date:[now-1day TO NOW] Note: Use date:[now/day-7days TO NOW/DAY+1DAY] for caching See https://cwiki.apache.org/confluence/display/solr/working+with+dates New Query Language Keywords TAG, SITE e.g. SITE:sales
Solr 1.4 v s Solr 4 Repository Details: Documents: 200,000 Folders: 5,000 Content Store: 4.77 GB Solr 1.4 Index: 2.99 GB Solr 4 Index: 2.03 GB Solr Content Store Cache: 749 MB (10-20%) Memory: No Alfresco cache for document join Much lower memory usage
Solr 4 Administration Solr 4 includes an Admin Interface Solr Install & Configuration Details Browse Schema Execute Queries Analyze Document Fields Inspect field values
Share Search Enhancements
Live Search Dynamic search results displayed as you type Results include: Documents Sites People Result previews with thumbnails Show more info One-click to view results
Filtered Results Results Filtered Simple, consumer-like search experience Support knowledge mining, discovery and re-use Features include: List and Gallery Results Views Quick Previews Contextual Actions
Filter Management Filter Management Page Control Where/When/How Filters are Display Supports Standard and Custom Properties
Search Usability Bookmarkable Search Page URLs Use as links, send in email.. Includes search terms and filters applied Term Suggestion As-you-type search term suggestions Spell-check Automatically suggest alternative spellings
Q&A Thank you