Squirro 3.2.4 - Release Notes

Owned by Saurabh Jain
Last updated: May 07, 2021 by Chiara Sannitz

This is the fourth minor release as the follow up to the last major Squirro release 3.2.0. This release was released on Feb 11, 2021 .

Contents - What's in the release?

New Features

  • Data Science Studio has now been renamed to AI Studio. All the APIs and user-interfaces have been updated to reflect that.

  • We offer a new type of Email template called Newsletters out of the box now. These email templates support GrapesJS editor out of the box to allow building richer Email templates. These newsletters complement our Cognitive Search offering as this allows us to send emails with latest results from subscribed communities as well as a list of trending communities.

Improvements

AI Studio

  • Performance improvements for fetching labelled data in AI studio.

  • Improve user feedback by communicating progress when creating many communities.

  • Ensure Ground Truth without any associated Candidate Set is displayed correctly.

  • Add ability to bulk add Labeled items to a Ground Truth.

  • Link Candidate set page to Ground Truth creation.

  • Link Ground truth labelling UI page to Build Model page.

  • Hide rules when proximity filters are disabled on the Ground Truth.

  • It is now possible to add an additional Candidate Set to a Ground Truth for an already existing ground truth.

  • AI Studio Models are not created anymore if the user cancels on the configuration screen mid-way.

  • AI Ground Truths are not created if the user cancels the configuration screen.

  • Focus view: Fixed Positioning for labels and skip button.

Cognitive Search Improvements

  • Cognitive Search project templates are now offered out of the box with a Squirro installation. A new Cognitive Search project can be created by choosing the right project template from the Create Project dialog.

  • On Communities set up screen, we now allow for re-import of facet values (if new values have been created ).

  • Cognitive Search: FOR YOU dashboard results are now sorted by date.

Further Improvements

  • Support query highlighting for entities fetching.

Regression Fixes

  • PDF importing broken due to trying to convert binary content to Unicode

  • We install poppler-utils when installing Squirro. Used for thumbnail generation.

  • Dashboard Editor shows now the project logo & name for Server Users.

  • AB App: Fixed a few FE regressions introduced with the last release.

Bug Fixes

  • 1-click connectors do not break anymore in the frontend if the connectors have any configuration options in getArguments method.

  • Sub-items are not lost when running KEE on them.

  • PDF importing no longer broken.

  • Fix duplicate highlight issues on PDFs.

  • Proximity based model should work again.

Breaking Changes

  • Favourite search functionality saves the favourite queries by default on a user level instead of the previous default of saving it on the project level. Also, we do not by default automatically create an email alert from the favourite search.

Fresh Installation Instructions

Please follow the regular installation steps

Upgrade Instructions

Please ensure that your current version is 3.1.0 or later. If you are on a version older than 3.1.0, please contact support.

With release 3.2.0 onwards, we have officially dropped the support for Python 2. If your custom Python plugins are not compatible with Python 3, please refrain from upgrading to any of the Squirro 3.2.x release until you have migrated the custom Python plugins to be Python 3 compatible.

yum update java-1.8.0-openjdk yum update squirro-storage-node-users # update storage node yum update squirro-storage-node yum update squirro-cluster-node-users yum update squirro-* # Resolve any `rpmnew` files. We anticipate `/etc/squirro/storage.ini`, `/etc/squirro/machinelearning.ini` and `/etc/nginx/conf.d/frontend.conf` to at least be resolved. systemctl reload nginx squirro_restart # Remove all orphaned python 2 packages yum erase squirro-python27*

Resolve all the `.rpmnew` files in `/etc/squirro`. This process involves merging the changes between the `.ini` (e.g., `storage.ini`) and '.ini.rpmnew`(e.g., `storage.ini.rpmnew`) files and then eventually deleting the `.ini.rpmnew` files. Finally, restart the services with merged `.ini` files (or just use `squirro_restart` to restart all services).

 

Upgrade the storage node by running:

yum update java-1.8.0-openjdk yum update squirro-storage-node-users # update storage node yum update squirro-storage-node

Upgrade the cluster node by running:

yum update java-1.8.0-openjdk yum update squirro-cluster-node-users yum update squirro-* # Resolve any `rpmnew` files. We anticipate `/etc/squirro/storage.ini`, `/etc/squirro/machinelearning.ini` and `/etc/nginx/conf.d/frontend.conf` to at least be resolved. systemctl reload nginx squirro_restart # Remove all orphaned python 2 packages yum erase squirro-python27*

Resolve all the `.rpmnew` files in `/etc/squirro`. This process involves merging the changes between the `.ini` (e.g., `storage.ini`) and '.ini.rpmnew`(e.g., `storage.ini.rpmnew`) files and then eventually deleting the `.ini.rpmnew` files. Finally, restart the services with merged `.ini` files (or just use `squirro_restart` to restart all services).

 

Upgrade every storage nodes (one by one) by running:

Upgrade each cluster nodes by running:

Resolve all the `.rpmnew` files in `/etc/squirro`. This process involves merging the changes between the `.ini` (e.g., `storage.ini`) and '.ini.rpmnew`(e.g., `storage.ini.rpmnew`) files and then eventually deleting the `.ini.rpmnew` files. Finally, restart the services with merged `.ini` files (or just use `squirro_restart` to restart all services).