Released 3rd May 2017

New Features

• Connecting to your data just got even easier: Data Loader Plugins - previously only available for use on the command line, can now be configured and used online from your Squirro installation. Just point and click...
• More Query Power: you can now use wildcards in Squirro facetted queries for example;
  • "company:*" ... match all items that have the company facet set
  • "-company:*" ... match all items where it's missing
  • "name:Jane*" match all items where facet "name" has a value that start with "Jane"
  • For further details see the Query Syntax documentation
• Steve, Steven or Stevie? It's all the same to us - we added support for synonyms allowing you to connect specific search terms which related to the same concept. Currently available via the Squirro REST API and Python SDK. Further details coming soon

Improvements

• Smarter Memory Use: we split off the Redis cache database from the rest of the the Redis databases, allowing us to use it's LRU caching capabilities, which will typically mean less memory usage overall. Worth noting is the documentation on flushing caches...

• Faster Service Restart: we optimised how Squirro services shutdown so less waiting around when restart Squirro or the server it is running on
• Catch Up with Facebook: the Facebook data provide now supports the latest version of Facebooks Graph API
• Flexible Tables: custom cell and header templates for Table widget
• Link with LinkedIn: improved detection of LinkedIn profiles
• Need to connect with HP Service Manager: we have DataLoader plugins for you...
• Ability to upgrade to the latest Java JDK security fixes in jdk 1.8.0 release 144 from July 2017 (added August 24, 2017)

Bug Fixes

• Fixed default provider config so that it works out of the box

• UTF-8 now properly handled in dataloader item templates

• Fixed various issues related to cache cleaning
• Fixed issue with squirro_asset tool and trailing directory slashes
• Fixed issue with facet mapping
• Fix for analyzer name starting with _
• Fixed issue with queries containing multiple spaces
• Fixed display of date in cards widget
• Fixed facet table detail overlapping
• Fixed throw on grid resize
• Fixed detail panel resizer opacity
• Fixed cleanup after source deletion
• Fixed a thread-safety issue with the query parser
• Fixed trend chart data grouping when using time selector
• Fixed KEE pipelets when uploaded via the squirro-loader using the --bulk-index option (added June 16, 2017)
• Fixed installation on systems with newer versions of python27 and dataloader stability fixes (added June 26, 2017)
• Fixed handling of optional files in squirro_data_load (added July 11, 2017)
• Fixed display of up-to-date items after adding keywords to existing items (added July 12, 2017)
• Updated python-nltk natural language processing support from 3.0.1 to 3.2.4. With that you can take advantage of ntlk's sentiment support in pipelets (added July 12, 2017)
• Fixed nginx installation issues in new installations on RedHat 6 or Centos 6 by moving to nginx-1.12-1 (added July 17, 2017)
• Fixed display of near-duplicates (added July 20, 2017)

• Allowing alert creation from dashboard search bar (added July 20, 2017)

• Allowing the squirro_data_load tool to enable near-duplicate-detection via the new --enable-near-duplicate-detection option when --bulk-index is not used (added July 21, 2017)

• No need to wait until ElasticSearch gets around to delete items after you remove a source. Now we clear out items associated with recently removed sources from the display immediately (added July 24, 2017)

• Robust filtering service SqlAlchemy session management (added September 5, 2017)
• Allow installing ImageMagick on Squirro cluster nodes (added August 28, 2017)