Connector Types

Database Connectors

Content stored in a database comes in the form of a set of rows returned in response to a query. Content is fetched via a series of database queries. The rows returned are transformed into documents for the index. The query results set returned by a fixed query will change over time. Database rows lack the kind of metadata information available on files, so tracking changes to documents is not always possible.

Fusion provides the following database connectors:

  • Couchbase which uses Couchbase’s XDCR feature to transfer data in real time

  • JDBC - for relational databases which can be accessed by a JDBC4 database driver, such as Oracle, MySQL, or Postgres

  • MongoDb

Filesystem Connectors

A filesystem-based data store is a network of nodes to be traversed, where each node (e.g., a Unix file directory) provides information about its child nodes (e.g., the files in that directory) or references other nodes (e.g. links in an html document). The extent of the network of nodes to be traversed is discovered during the crawl. The crawler captures information about the node, e.g., file name, permissions, date of creation, last modification, and last access, as well as the contents of the nodes.

Fusion provides the following filesystem connectors:

Hadoop Cluster Connectors

Fusion provides connectors for the following Hadoop distributions:

  • Apache 2.x

  • Cloudera

  • Hortonworks HDP

  • MapR

  • Pivotal

See section Hadoop Connector and Datasource Configuration for connector and datasource details.

Logging Connectors

  • Logstash (deprecated) - collects logging and event streams from a Logstash instance

Push Content Connectors

  • Push connector- push content to Fusion for indexing into Solr via a Fusion index pipeline

Script Connectors

Social Media Connectors

Web Connectors

WebSphere Connector

  • Websphere - The Fusion WebSphere connector requires Solr. Configuration and integration is more complicated than the other Fusion connectors.