Skip to main content

Cogynt v2.9 Release Notes

· 9 min read

Cogility is excited to announce that Cogynt version 2.9 will be generally available for both AWS and GCP customers as of April 2023. Please see what's new below.

If you'd like more information, or want to have your Cogynt instance updated, please contact us.

Authoring​

Authoring Experience Improvements​

  • Users are now required to specify when partitions are not desired in a pattern. This will prevent users from deploying incomplete patterns with missing partitions, which can potentially cause serious data corruption.
  • Partials support has been removed in favor of using multiplicity.
  • These changes help Data Scientists and SMEs follow best practices and catch mistakes earlier in the modeling phase.

Risk Filter​

  • A risk threshold can be applied to patterns with risk analysis implemented, reducing bloat from unwanted or unimportant data.
  • The filter adds a new dimension and increased flexibility and control in risk analysis.

Optimized Performance​

  • Improved array serialization to boost array processing efficiency in Flink jobs.
  • Reduced project and modeling component export file size and improved export/import speed.
  • Optimized Flink state storage by removing unnecessary storage of unused fields and metadata, which in turn boosts checkpointing performance.

Deployment Configuration UI Upgrade​

  • Flink configurations are now displayed with user-friendly and intuitive labels.
  • Tooltips clearly describe the purpose of each configuration, and provide guidance on what input values to type in.

Other User Experience Improvements and Bug Fixes​

  • When viewing the lexicon, right-clicking opens an actions menu.
  • The deployment stop button has been updated for consistency with industry standards, making buttons more recognizable.
  • Improved deployment status UI accuracy and stability.
  • More improvements and bug fixes have been added to improve the authoring experience.

Data Management Tool​

S3 Data Connectors for CSV Files​

  • Data Scientists and SMEs can now create and manage CSV to S3 connectors through a dedicated UI and tool in the Cogynt Data Management Tool.
  • Users can start, pause, stop, and reuse connectors to manage their data pipeline from S3 buckets.
  • Users can now map fields from CSV headers or columns to a target field in a user schema from a project.
  • An added layer of data validation ensures preservation of data integrity.
  • Previously, users established this connection via Confluent Control Center, where a JSON file is configured and uploaded manually. Having a dedicated UI to help build and manage connectors improves the usability of the custom CSV to S3 connectors.

Optimized Performance​

  • Support multi-byte character in CSV S3 connector.

Support for JSON Files in an S3 Location​

Note

This feature is only available in Confluent Control Center.

  • Data Scientists and Data Engineers may establish a connector for JSON files stored in AWS S3.
  • These users establish the connector for the Confluent Control Center.
  • This allows Data Engineers to make JSON files available to Authoring Tool users on a production scale.

User Management​

  • The "super admin" role now appears on the list of roles as a "read only" role.
  • Any current "super admin" can now assign or revoke this role for additional users.
  • In the rare case that the authentication provider is changed for a cluster, the first user that logs in using the new authentication provider is automatically delegated as the "super admin."
  • The changes allow users more control over the "super admin" role and prevent scenarios where the "super admin" role is locked out.

Workstation​

Major Improvements to Collections Widget​

The Cogility Team has re-imagined Collection Comments and Custom Fields. These benefits make it easier to collaborate and write lengthy, detailed reports.

Comment Improvements​

  • Collection Notes are retitled as "Comments."
  • The comments user interface is now easier to use and digest.

Custom Field Improvements​

  • Analysts may now edit an individual field just by clicking on it, rather than by clicking an Edit icon that changes every field to editable.
  • The UIs of all custom fields have been reworked.
  • "Rich Text" fields are now available:
    • Allows analysts to write out information with up to 50,000 characters.
    • Analysts may add style and formatting to their text.
    • These fields are exportable to PDF using the Report Builder templates.

Events Stream Widget User Experience Improvements​

  • The UI now hides the Core ID field and adds a Reset button that hides manually toggled fields.
  • This improves the analyst's ability to investigate events by surfacing only the most relevant fields.

Event Detail Templates are Manageable by All Analysts​

  • Any user with access to Designer > Event Detail Templates may view any event detail template. Previously, only the author of a template could manage it.
  • Analysts can leverage all available templates to better understand the underlying details of an event.

Notifications Explorer Widget Loading Message Improvement​

  • Created a new loading message for the Notifications Explorer.
  • It is now clearer to the analyst when events are simply loading, instead of no events being returned.
  • Spacing slider functionality has been improved.
  • Retention to link chart settings has been added.
  • These improvements allow analysts to better control the display of knowledge graphs and retain their settings.

Other Bug Fixes and Improvements​

  • Custom Field single selection and checkbox fields with many options are now scrollable.
  • Switched map provider, resulting in increased performance and stability.

Audit Viewer​

Several new actions are now logged or have been enhanced in the Audit Viewer and the _cogynt_audit_log Kafka topic. For more information, see Logged Actions in the Audit Viewer User Guide.

Authoring​

  • Enhanced and additional logs for Flink deployments.
  • Logging for projects.
  • Logging for user data schemas.
  • Logging for event types.

Workstation​

  • Logging for applied search facet filters.
  • Logging for Custom Field Templates.
  • Logging for Event Detail Templates
  • Logging for Report Builder templates.
  • Logging for custom fields on a collection.

Data Management​

  • Logging for Kafka connectors.
  • Logging for manually entered data.

DevOps​

Increased Observability​

  • New observability suite with S3 compatible storage.

Increased Security​

  • New in-cluster security suite using OPA Policy-as-Code.

Other Improvements​

  • PostgreSQL WAL performance improvements.
  • Pinot Deep Storage support for S3.

Known Issues​

Authoring​

  • For GCP clusters, the Flink Operator pod may fall into a restarting cycle, and cause the start deployment action to stall until the Flink Operator becomes active again.
  • Geo-polygon constraints with geo-coordinates will throw Flink errors.
  • Applying the toString function to a non-generic array will cause deployment to fail. If this is required, change the array to a generic array.
  • The deployment target edit screen may sometimes be locked upon first opening a new project. To circumvent this, exit out of the deployment target's edit screen and re-enter the edit screen.
  • "Something went wrong" error toast may appear upon deleting an event pattern. This error banner can be ignored.
  • Locking behavior may sometimes behave erratically, where objects may be locked when no other user is actively working on the object.
  • Pattern layout is not preserved upon project export. Use auto-layout to manage layout in the meantime.

Data Management Tool​

  • Using the tab delimiter in the CSV S3 connector may result in null fields.
  • "Something went wrong" toast may appear when first entering a new connector edit page.
  • Create & Run and Update & Run buttons are unresponsive upon creating or updating.
  • Creating a new topic in the S3 connector configurations screen will not auto-select the new topic if another topic was already selected previously.
  • Inaccurate filter results when filtering by multiple status.
  • File upload preview fails to show undefined fields when Event Type field is missing in CSV or JSON file.

Workstation​

  • The Events Stream widget does not properly sort columns that are named with camel-case format or with spaced words.
  • In rare circumstances, adding attachments to a comment may require the user to navigate to another tab (like Custom Fields) and back to the Collection tab to see the change.
  • Deleting an attachment on a reply comment requires the user to navigate to another tab (like Custom Fields) and back to the Collection tab to see the change.
  • In some circumstances, a retracted pattern solution on the Drilldown widget may change color between blue and gray. Retracted pattern solutions are intended to always be gray.
  • There are rare situations in which the Drilldown widget may show outdated information for an event. If this happens, a reset should be performed to force the reingestion of Drilldown data.
  • The Expand Group feature of Link Analysis does not work for manually collapsed nodes on a link graph.
  • The tag filters on the Notifications widget do not accurately show the count of system tags.
  • The interface for adding hyperlinks to a rich-text custom field may appear cut off if the target text is to the left of the text input area.
  • A user that is assigned to an event notification will not receive a system notification when the event notification is deleted by ingestion.

User Management​

  • Deleting a user may cause collections to fail to load on the Collections widget. It is not recommended to delete users at this time.

Audit Viewer​

  • Updates to a collection (name, description, priority, tags, assignee, etc.) may not be logged in Audit Viewer when there are a large amount of custom field templates created. The logs will exist in Kafka, however, which allows for a recovery in a future release.
  • A deleted event detail template does not include the deleted_at timestamp.
  • Only in GCP environments: Sorting by column headers produces an error.

Other​

  • Cogynt clusters with a license in the grace period incorrectly prevents users from accessing User Management and Audit Viewer.