Release Notes

For changes in version 4.0 and above, see the Fusion Server release notes.

Note
Only Fusion versions 3.0.0 and above are currently supported.

Current Fusion Release: 3.1.5

Release date: 13 April 2018

Component versions:

  • Solr 6.6.3

  • ZooKeeper 3.4.6

  • Spark 2.1.1

Major changes:

  • Apache Ignite has been updated to 2.3.0.

Fusion Release History

Version Release Date Major Changes Component Versions Release Notes

3.1.4

29 March 2018

  • Solr 6.6.3

    This release incorporates Solr 6.6.3, which includes the following bugfixes:

    • Disallow reference to external resources in the DataImportHandler dataConfig request parameter

    • Allow collections created with legacyCloud=true to be opened if legacyCloud=false

    • LeaderInitiatedRecoveryThread now retries on UnknownHostException to better handle changing IP and restarting Fusion

      Additional changes are described in the Solr 6.6.3 release notes.

  • Sharepoint connector improvements

    • Fixed XML error in ADFS SAML authentication.

    • Improved logging to help troubleshoot document onboarding.

    • Fixed a bug in which the online instance was not downloading body content.

  • Zendesk connector improvements

    • Now crawls knowledge base articles in the Zendesk Help Center.

    • Zendesk security filtering now supports access control lists.

  • Web connector improvements

    • The Web connector can now be configured to retry before failing a request.

      To use, set the new API parameter f.requestRetryCount to a value other than the default 0.

    • Web sites protected by Ping Identity can now be authenticated with the Web connector.

    • Can now add HTTP response header fields to an index.

    • Can now normalize URIs. Use the httpClient parameter URIUtils.resolve to resolve the URL if needed.

  • Other improvements

    • Regex field replacement now correctly does the following:

      • Removes double-square brackets around wiki references

      • Removes XML comments

      • Checks that the regex replacement is not empty

    • Improved page load performance during the JavaScript evaluation stage using the JavaScript connector.

    • Form login now works with login pages that have input tags, instead of only form tags.

    • Windows Share crawler max_bytes property can now handle large files.

    • Significant performance improvement in the Jobs UI page, at Collections > Jobs.

      We’ve introduced a system job to clean up old entries in job history.

    • Proxy socket timeout is now configurable.

      In fusion.properties, add -Dcom.lucidworks.apollo.admin.proxy.socket.timeout= to proxy.jvmOptions.

    • The Service Now connector now supports communication through a proxy.

  • Solr partial indexer stage failure error changed from 204 to 5XX.

  • Minor usability improvements in parser configuration UI.

  • Fusion UI now validates the name field for an aggregation job.

  • Solr 6.6.3

  • ZooKeeper 3.4.6

  • Spark 2.1.1

3.1.3

19 December 2017

  • The Web connector now crawls JavaScript-enabled Web sites faster with the use of a headless instance of Firefox as a crawling agent.

  • Jive events, videos, and ideas can now be indexed. A new fetch_personal_blogs parameter enables fetching user blogs.

  • The default value for the Jive connector places_to_crawl has changed from "all" to "space,group".

  • The Jive connector now stores and retries documents without ACLs. It also updates the index when ACLs have changed.

  • The Sharepoint connector no longer retries failed crawls by default.

  • Solr 6.6.2

  • ZooKeeper 3.4.6

  • Spark 2.1.1

3.1.2

24 August 2017

  • The Web connector now supports links in RSS feeds by extracting links from <link> elements.

  • This release fixes a known issue in Fusion 3.1.1 by correctly hiding passwords while they are being entered in object configuration form fields in the Fusion UI.

  • When time-based partitioning is configured for the searchlogs collection, Fusion now correctly partitions the data. This fixes a known issue with Fusion releases between 2.4.0 and 3.1.1.

  • When the MongoDB connector encounters multiple values for the same field, it now copies all values to that field, as a list.

  • The Jive connector now retrieves access control lists (ACLs) for blog posts whose parent is a person.

  • Solr 6.5.1

  • ZooKeeper 3.4.6

  • Spark 2.1.1

3.1.1

17 August 2017

  • All Fusion services can now be SSL-enabled.

  • Query-pipeline stages: default scale ranges added for recommendation boosting with signals; Solr query parameters for Boost with Signals are configurable in the UI.

  • Google Drive connector: exports to Excel spreadsheets; has configuration parameters for the complete set of search clauses for the Google Drive API file.list method; improved crawling robustness

  • Jive connector: now crawls people and profile pages, and correctly retrieves all batches.

  • Apache Tika parser: added configuration parameters maxDepth and maxPackageEntryDepth

  • Fusion SQL: timestamp ranges in queries are sent to Solr automatically.

  • Windows Share connector: allows guest users.

  • Catalog API: support added for field aliases.

  • Spark jobs: additional Spark job subtypes were added – Co-occurrence Similarity, Item Similarity Recommender, Levenshtein, Collection Analysis, Statistically Interesting Phrases (SIP), Doc Clustering, Outlier Detection, and Cluster Labeling.

  • Solr 6.5.1

  • ZooKeeper 3.4.6

  • Spark 2.1.1

3.1.0

20 June 2017

  • Fusion now ships with a set of basic connectors, reducing the download size. Additional connectors can be downloaded separately.

  • Fusion now fully supports machine learning

  • Fusion’s recommendations are now powered by Spark, with new query pipeline stages and a new recommendation jobs interface.

  • The Fusion UI now has Object Explorer, a means of visualizing Fusion objects and their relationships.

  • The new Sharepoint Online connector supports crawling Sharepoint Office 365 online sites.

  • New parsers – Fusion now has an HTML parser and an XML parser.

  • New APIs – Fusion now has Links, Jobs, Scheduler, Tasks, and Groups APIs.

  • You can configure Fusion and an external SolrCloud cluster so that communication between Fusion and the SolrCloud cluster is secured.

  • Jobs and schedules are now separate. A new Jobs interface lets you define tasks and Spark jobs and schedule all types of jobs. The Scheduler interface has been redesigned.

  • Blob store management has been improved.

  • Index pipelines for connectors are no longer embedded in the connector nodes (by default). You can scale and load-balance indexing.

  • The Jira connector has improved fetching performance, and supports parsers and security trimming.

  • The Web connector has a new f.crawlJS that lets the connector crawl JavaScript-powered websites accurately.

  • The Jive connector now parses lists and map data.

  • The Box.com connector has better support for incremental crawling.

  • Solr 6.5.1

  • ZooKeeper 3.4.6

  • Spark 2.1.1

3.0.1

4 April 2017

  • Support for Sharepoint 2016

  • Box.com connector: support for proxying and security trimming

  • Google Drive connector: support for security trimming, enterprise crawl, and Google native formats

  • Improvements to the MongoDB connector, Solr Partial Index Updater index stage, Advanced Boosting query stage, and Active Directory Security Trimming query stage

  • Simplified upgrade procedure

  • Solr 6.4.2

  • ZooKeeper 3.4.6

  • Spark 1.6.3

2.4.5

20 March 2016

  • Support for Sharepoint 2016

  • Box.com connector: support for proxying and security trimming

  • Google Drive connector: support for security trimming, enterprise crawl, and Google native formats

  • Improvements to the MongoDB connector, Solr Partial Index Updater index stage, Advanced Boosting query stage, and Active Directory Security Trimming query stage

  • Solr 5.5.4

  • ZooKeeper 3.4.8

  • Spark 1.6.1

3.0.0

25 January 2017

  • Launcher

  • Quickstart UI

  • Query Workbench

  • Parsers

  • New index pipeline stages

  • New query pipeline stages

  • Improved startup scripts

  • Support for desktop uploads

  • New APIs

  • Fusion services as Windows services

  • New user roles

  • Memory footprint reduction

  • Improved Index Workbench

  • Improvements to the Hadoop connector

  • Aggregation jobs enabled by default

  • Support for Jive groups

  • Support for Amazon SSE-KMS

  • Better OAuth integration

  • Support for XML in the Index Pipelines API

  • Solr 6.3.0

  • ZooKeeper 3.4.6

  • Spark 1.6.3