Changelog

v0.20.0

Released 2025-11-20

New Features

• Field Description Values

  • Add reference values for Field Descriptions. The values
  • There is a new “values” tab on the Field Description page for Field Descriptions that have the role “Dimension”. This tab lists all field description values and has the option to add, edit and delete values.
  • Field Descriptions with values can be expanded in the Field Descriptions overview table to show the individual values as children in the table.

• Team Ownership

  • Assign user groups as owner to assets or to custom roles through custom attributes of type “user”.
  • Assigned assets for the user group is visible in the user group page.
  • Assets assigned to users through a user group will show up on the user’s profile page, and a new column shows if the ownership is direct or through a user group. Hovering on the role or label displays a tooltip to show which user group the assignment is inferred from.
  • There is a new option when creating user groups to only sync user group memberships or also add/edit/delete users to Katalogue that are members of the user group. This enables the case to have a few Microsoft Entra Id user groups that control access to and role in Katalogue, and other groups that can be used as teams to assign assets to.

Improvements And Changes

• Not NULL Attribute for Fields/Columns Showed on the Dataset Page.

  • This was previously only shown in the Field/Column details pane.

• Re-designed Search Results Page

  • Asset type categories on the main search result page are now vertically aligned to allow more categories and a better user experience. They were previously horizontally aligned above the actual search results.

• User Provisioning Improvements

  • BREAKING CHANGES. This is an important change in user management behaviour - read the User Management section carefully!
  • The purpose of this change is to fully automate user management when integrating with Microsoft Entra Id and to enable the team ownership feature described above.
  • A summary is that user groups now can be defined as “user provisioning”. Such user groups change user roles and also add/delete users in Katalogue depending on the user groups a user is member of. Non-user provisioning user groups only sync membership with existing users in Katalogue.
  • When upgrading to this Katalogue version, all existing Microsoft Entra Id user groups will be defined as “user provisioning”. Local user groups will not.

• Table cell content is top-aligned

  • Previously, content in all cells were vertically centered, which made it hard to visually scan tables with line breaks on a row.

• Long Descriptions are Truncated

  • Long descriptions in table views are truncated.
  • Long descriptions in the details pane for an asset is truncated by default, but can be shown in full by a button click (only shown if the description is truncated).
  • The purpose is to enhance the UX for extremely long descriptions.

• Changelog for Microsoft Entra Id sync tasks

  • The changelog related to user groups and users for Microsoft Entra Id sync tasks is now visible in the GUI for admins. This allows more insights in exactly which user-related changes were made in a sync job.

Bug Fixes

• Fixed: Navigating to a another dataset directly from the current dataset view (i.e. by clicking on a search result) and then interacting with the table component throws the user back to the previous dataset.
• Fixed: A few minor GUI glitches

v0.19.0

Released 2025-10-06

New Features

• New GUI design with dark mode support

  • The entire GUI has been redesigned for a number of reasons:
    • Reduce clutter to give an even more minimalistic appearance
    • Give the GUI an updated, more modern look and feel
    • Improve contrast of some elements to improve usability
    • Refactor the code base to allow for future development
  • This redesign also enabled theming with dark mode support
    • Select color theme in the User Profile menu in the upper right corner
  • The redesign also includes many minor GUI improvements, too many to list here.

• Katalogue REST API

  • The main use case is to enable scripts and other applications like dbt and Airflow to interact with Katalogue.
  • All assets in Katalogue have CRUD (Create, Read, Update, Delete) endpoints which enables programatical interaction.
  • Complete OAuth2 authentication mechanism which supports the client credentials flow. Clients can be registered in the GUI (Settings -> REST API)
  • Endpoint authorization uses the OAuth2 “scope” mechanism.
  • Swagger documentation for endpoints, which is accessible at /swagger. Note that the swagger documentation is currently not complete.
  • Most aspects of the REST API is configurable, such as the access token and client secret TTLs, token endpoint and swagger paths. See the docs for details
  • Supports external private keys for access token signing. By default, an internally generated private key is used.

• Katalogue REST API system and glossary export endpoints

  • It is now possible to easily export entire systems and glossaries
  • The export contains all child assets, e.g. for systems: datasources, dataset groups, datasets and fields, including any assigned custom attributes.

• Customizeable data classification labels

  • Easily implement the data classification nomenclature from your organization.
  • See Settings -> Enumerations -> Data Classification

Improvements And Changes

• New Enumerations section on the Settings page
  • The “Field Units” page is moved here from the main navigation menu
  • New “Datatype” page to manage datatypes through the GUI
  • The previously mentioned “Data Classification” page is also organized in this section
• Field attributes like precision, scale, length, is nullable and default value is exposed in the add/edit field dialog to allow manual edit of them.
• User Group photos can be synced from Microsoft Entra ID. It seems Microsoft has solved the bug in the Microsoft Graph API that prevented this earlier.

Bug Fixes

• N/A

v0.18.0

Released 2025-08-18

New Features

• Business Term Hierarchies

  • Enables nesting of business terms to create hierarchies in glossaries.
  • Edit a Business Term and set the new attribute “Parent Business Term” to another, already existing business term in the glossary. This populates the new column public.business_term.parent_business_term_id in the database.

• Include Field Descriptions in glossaries

  • Field Descriptions that refer to (has an @mention reference to a business term in either the Description, Example or Definition textboxes), can now be included as children to the business term in the glossary view. This further enhances the usage of Field Descriptions as attributes to business terms and is an important step towards representing information models as a Katalogue glossary.
  • Enable this through the new option “Include Field Descriptions” in the toolbar next to the filter function in the business terms table in a glossary.

• Field Hierarchies

  • Enables representation of nested datasets like JSON structures.
  • Edit a Field and set the new attribute “Parent Field” to another, already existing field in the dataset. This populates the new column public.field.parent_field_id in the database.
  • Automatically syncing nested datastructures automatically from datasources, like the Databricks struct datatype, is not yet supported.

Improvements And Changes

• Minor GUI polish.

Bug Fixes

• N/A

v0.17.0

Released 2025-06-19

New Features

• Custom Attributes
  • Enrich assets in Katalogue with custom metadata attributes.
  • Define custom attributes and see all assets that have a specific custom attribute in the new Custom Attributes page. This page is available for admins in the “Manage” section of the navigation pane.
  • There are six different types of attributes to chose from. The types define how they will be showed in the GUI:
    • String: Short text, displayed as a one-line input
    • Description: Long text, displayed as a textarea with formatting enabled
    • Code: Long text, displayed as a textarea with monotype font and syntax highlighting
    • Boolean: True/false values, displayed as a checkbox
    • URL: Link to an external resource, displayed as a one-line input with the option to add a display name for the link
    • User: Reference to a Katalogue user, displayed as a dropdown menu with existing Katalogue users to chose from. Can be used to link additional users with other roles than “Owner” to assets. These assets will also show up in the User and Profile pages along with the assets the user is an owner for.
  • Custom attributes are currently not part of the search index, meaning that the custom attribute values are not searchable. This will be changed in a future release.

Improvements And Changes

• Re-designed Owner column in tables. The owner name has been removed and the column displays only the photo or an icon with initials if no user photo is present. Previously, the full name was displayed. This change makes the column narrower to give more place for other, more relevant columns.
• Environment label. Set the Katalogue instance environment in Settings -> General. This makes it easier to visually distinguish between Katalogue instances and avoid confusion when working with multiple Katalogue instances in the browser. When defined and when the label is not “PROD”, this label is visible in the GUI, the browser tab and in emails. This label can be retrieved from the database with the following query: SELECT setting_value FROM public.setting WHERE setting_code = 'ENVIRONMENT'
• New attribute on Systems: system_code. This attribute can be used as an alias or internal ID for a system.
• IBM DB2 connector improvement. Native support for z/OS platform. Previously, the connector only had a native SQL query against iSeries/AS400, and connecting to z/OS required custom sql queries. It is now possible to select platform in the connection and the corresponding query will be used.
• Support for setting all config parameters that are not present in the database table public.settings as environment variables see docs. This makes deployment easier.
• Creating a Connection will by default also create a Datasource Sync task for it. This makes it a bit faster to setup a new connection as the step to manually create a Datasource Sync task is now automated. Disabling this can be controlled with a new checkbox, which is checked by default, in the Add Connection dialog.

Bug Fixes

• Lots of bug fixes and stability improvements since the last version…

v0.16.0

Released 2025-02-22

New Features

• Dataset Lineage
  • View lineage on dataset level on the new “lineage” tab for datasets
  • Automatically syncs lineage from datasources that provide lineage (currently Databricks and dbt)
  • Ability to add a custom lineage SQL query to import lineage from other datasources. Use this to e.g. ingest lineage from a table with pre-calculated lineage or provide your own rule-based lineage logic such as “source_table_name = ‘src_’ + target_table_name”
• dbt Connector
  • Upload a local manifest.json file or connect to a live dbt docs URL
  • The dbt connector can be combined with other connectors to blend metadata from multiple sources, or used stand-alone.
  • This initial release supports import of the following metadata:
    • schema (such as tables, columns, datatypes etc)
    • descriptions
    • lineage
    • code / technical definition (both raw and compiled)
  • Not yet supported:
    • tags
    • meta
    • owner
    • relationships
    • exposures
• Azure Key Vault integration. Integrate with Azure Key Vault to let Connections use passwords stored there instead of storing the password in Katalogue.
• Automatically sync description changes from the datasource There is a new option for fields, datasets, dataset groups and datasources to select source for the description. The source can be set to “katalogue”, which means the description will be static, or to the source system, which means it will update if the description in the source updates.

Improvements And Changes

• GUI improvements
  • The width of the asset property pane and add/edit assets dialog pane can now be adjusted
  • Technical attributes in the asset property pane are grouped in a new details section that can be expanded/collapsed
  • Improvements to following the status of running jobs in the Task and Job views. Most notably is that the “runtime” attribute updates in real time to show actual elapsed time of the job. There is also a new “warning” status to highlight minor non-stopping encountered during the job execution.
  • Re-designed dataset popup table when selecting a dataset in the relationship graph
    • It is moveable
    • It is resizeable
    • The content uses available space better
    • There will be one popup table for each selected dataset
• Support for setting some config parameters as environment variables see docs. This makes deployment easier.
• Support for passing sensitive config parameters as Docker Secrets see docs. This makes deployment easier and more secure (no secrets are stores in the Docker container)
• Datasource connection settings has been moved from the datasource object into a new Connection object. This gives a clearer structure and is a step towards setting up one connection that scans all databases on the same host. Currently, there must be one connection for each database.
• Major rewrite of the datasource sync functionality BREAKING CHANGES see below! The entire datasource sync logic has been rewritten to achieve a number of improvements:
  • BREAKING CHANGE. More logical property names. When using custom import SQL queries, the property “field_description_description” for column descriptions has been renamed to “field_description”. All custom SQL queries must be updated with this new name, otherwise the datasource sync task will fail.
  • When using custom import SQL queries, the dataset_type property is now case-insensitive. Previously, the property had to be exactly “Table” or “View”.
  • Columns with datatypes unknown to Katalogue are automatically added and synced. Previous behaviour was that if a datatype was unknown to Katalogue (i.e. not present in public.datatype table), all columns with that datatype would be silently ignored when syncing with a datasource. New behaviour is that datatypes not present in the public.datatype table will be automatically added and thus columns with those datatypes will be synced as expected. However, those columns will show up with a “unknown datatype” icon in the GUI. To fix this, the public.datatype.datatype_category value must be manually changed from “unknown” to some of the other valid values for the property.
  • Fields, Datasets, Dataset Groups and Datasources are now primarily matched (case-insensitively) on name and secondary on source id (if available in the source, see documentation) when syncing datasources. This enables Katalogue to handle name changes with an update instead of a delete & insert. Previous behaviour was to only sync on name and therefore not handle name changes in a good way. This change also enables manual creation of assets in Katalogue and that later will be automatically synced when added to the underlying datasource.
  • Change: Field Descriptions are no longer automatically created, they must from now on be created manually in Katalogue
• Column properties “Is Nullable”, “Ordinal Position” and “Column Default” synced These column properties are now automatically synced from relevant datasources. They are shown in the GUI in the “details” section for Fields.
• Databricks connector improvements. The Databricks connector now supports all standard connector features like datasource filters, custom sql queries etc.

Bug Fixes

• Fixed: When navigating to e.g. a field description from the field page and deleting the field description, the user is unexpectedly returned to the list of field descriptions instead of the field page.
• Fixed: Search strings containing ”(”, ”)” and a few other characters crashed the backend api. Invalid search strings are now handled gracefully and does not crash the app.

v0.15.0

Released 2024-03-28

New Features

• ODBC connector. Support to use any ODBC connector to connect to datasources that does not have a native connector, such as Progress OpenEdge.

Improvements And Changes

• Support for multi-column relationships. Multi-column relationships are now imported and shown correctly in the GUI and relationship graphs.
• Renamed “References” tabs to “Related Assets”. This is to better reflect what the tab contain. A related asset is an asset such as a dataset that is referencing/“mentioning” the first asset in a description field. Example: Dataset “customers” has the description “Table where each row represents a @customer” and where “@customer” references the business term “Customer”. The dataset “customers” will then show up as a related asset to the business term “Customer”.
• Connector Settings in GUI For admins, there is a new tab “Connectors” on the Settings page where some settings related to connectors can be configured directly in the GUI. These were previously located in the backend api service config file “appsettings.json”.
• Ability to disable connectors It is possible to completely disable specific connectors. This is mainly useful in some test- and development scenarios. See the documentation in /spa/README.md for more info on this feature.

Bug Fixes

• Fixed: The relationship graph and dataset popup window received unnecessary scrollbars in some occasions
• Fixed: The frontend GUI crashes when quickly navigating to the relationship graph tab, i.e. before the dataset information has been loaded.

v0.14.0

Released 2024-02-22

New Features

• IBM DB2 connector. IBM DB2 for i-series and z/OS is supported. z/OS requires custom import query to work, see below.
• Custom Import Queries. Write your own custom SQL to import metadata from DB sources. This feature is a replacement/extension of the previous “use custom import view/table” feature. In the custom query editor, it is also possible to insert the default query to review and modify it.
• Syntax highlighting for SQL code. SQL code has syntax highlighting in every place in Katalogue where you can see SQL, e.g the Technical Definition tab for Views (datasets). Syntax highlighting is also applied to “view query” buttons in the add/edit datasource dialog.
• Show descriptions in table views. Tables in the GUI has a new button to toggle between showing statistics (documentation level, number of datasets etc) and the description of an asset.

Improvements And Changes

• Performance improvements:
  • Editing and deleting assets is now fast. Previously, saving a datasource could take up to several minutes. This was related to the search index update, which takes some time if Katalogue contains a lot of assets. Search index updates are now performed in the background, meaning that it can take a couple of minutes before e.g. a name change becomes searchable.
  • Loading Jobs in a Task is now much faster
  • The “update search index” step in datasource sync tasks is much faster
• GUI improvements:
  • Current page name is shown in the browser tab (previously, the tab showed “katalogue” regardless of page)
• Connector improvements:
  • MSSQL connector now includes the “MS_Description” extended property as field and dataset description. The MSSQL connector did previously not support import of descriptions/comments from the database.

Bug Fixes

• Fixed: Oracle connector had a bug that on some occasions produced wrong Primary Key information for columns.
• Fixed: Changelog information does not show if the changelog table for some reason misses old rows.
• Fixed: Reset password bugs:
  • Users get a Reset password token expired error when trying to reset their password. This was related to a timezone mismatch issue.
  • Link in reset password mail is wrong, which causes the user to reach the login page instead of the reset password page (the URL contained an extra /)
• Fixed: Updating assets with 10000+ related assets failed
• Fixed: A few GUI glitches resolved.

v0.13.0

Released 2023-10-24

New Features

• Databricks connector (requires Unity Catalog)
• Select which column(s) to filter in table filters
• Changelog for data assets shown in GUI
  • New tab on a data asset: Changelog
  • See all changes that ever happened to an asset (available for Systems, Datasources, Dataset Groups, Datasets, Fields, Field Descriptions, Glossaries, Business Terms and Users)
  • It is possible to filter the changelog
• Changelog for datasource sync tasks
  • See all changes made by a datasource sync task.
• Schema change notifications
  • New field in add/edit task dialog to enter a list of email addresses that will be notified if the sync task results in schema changes (changed names, changed datatypes, new/changed/deleted columns in datasets etc)
  • Any Katalogue user can view the schema changes, even Viewers.
• Job steps for a task job shown in GUI. Including runtime for each step.
• Tooltips on column headers

Improvements And Changes

• Minor GUI improvements:
  • The Description field in the datasource add/edit form is moved to the General tab to give more space and prevent scrolling.
  • Katalogue logo is moved to the navigation menu to give more space for breadcrumbs
  • Added hostname information on the datasource details pane
  • Added PII and data sensitivity information in search results
  • When minimizing the data asset details pane, it stays minimized when navigating to other pages (previously, it was maximized when visiting new assets)
• Improved “custom information schema view” experience.
  • Better error messages when the view does not fulfill the requirements
  • Basic data cleaning to convert empty strings to null.
• Connector improvements:
  • MSSQL now support connecting to named instances

Bug Fixes

• Fixed: Sync tasks for large datasources would sometimes hang on the “upserting relationships” step and never compelte
• Fixed: Navigation between tabs, and back and forth between data assets is more stable. Previously, on some rare occasions, the filter state, selected tab etc would be lost when navigating back to a previously visited page.

v0.12.0

Released 2023-07-07

New Features

• Fields now has enriched datatype information with character length, numeric precision and numeric scale.

  • This information is visible in the tooltip for datatype icons and as a new column in the field-list in a dataset.
  • The information can also be retrieved directly from the database by using the UDF to_datatype() in a select clause, see example below.
  • There are exceptions for some datatypes. Special datatypes like “interval” will not have this information. Datatypes like “timestamp” willnot be shown with the fractional second precision in the GUI, nor by the to_datatype() UDF, eventhough that information exists in the fields.datatype_precision column.

  -- The column datatype_fullname in this query returns "varchar2(32)", "decimal(32,2)" etc
  SELECT
      f.field_name
      , to_datatype(
            dt.datatype_name
          , dt.datatype_category
          , f.datatype_length
          , f.datatype_precision
          , f.datatype_scale
        )                         AS datatype_fullname
  FROM public.field f
  INNER JOIN public.datatype dt ON f.datatype_id = dt.datatype_id

• Statistical information like number of rows and total disk size (including indexes) are synced for database table datasets.

  • Row count are shown as a new column in the dataset-list in a dataset group
  • Both row count and disk size are shown in the dataset details pane (if the information is not null)
  • Both row count and disk size are available as node size measures in the relationship graphs.
  • Both row count and disk size are saved in the new public.dataset_statistics table for direct SQL integration.
  • The row count comes from the database statistics tables, which means that the numbers are an estimation and not an exact count.
  • For Oracle datasources, read permission on the DBA views dba_segments and dba_indexes is required to sync disk size information.
  • For PostgreSQL datasources, row count is only available on tables that have been VACCUM ANALYZEd.

Improvements And Changes

• It was previously not possible to sync oracle datasources for users without read permissions on the v$database (or v$pdbs for PDB databases) view. They normally require DBA permissions to query. This is now possible but metadata like database creation date is not synced without read access to those views.

Bug Fixes

• Fixed: Default sort order in tables are on some systems case-sensitive, meaning that PRODUCT was sorted before customer. The issue was related to the underlying OS implementation of various collations

v0.11.3

Released 2023-06-12

New Features

• Log all incoming HTTP requests to see origin ip, username, runtime etc for each API request. This feature can be enabled in the Settings->Logging section

Improvements And Changes

• Improved performance for search index updates (runs x3 times faster)
• Improved performance for saving field descriptions (runs x100 faster)
• Improved performance for datasource sync tasks (runs x10 faster for datasources with 100 000+ columns)
• Task last run timestamp and runtime is updated in the GUI when a task completes (previously, only task status was updated, all other info required a page reload)
• Track changes performed by a datasource sync task. A new column, transaction_id, is added to the job and changelog tables. Join the changelog table to the job table on this column to see exactly what was changed by a datasource sync task.

v0.11.1

Released 2023-05-02

New Features

• Improved search functionality
  • Create more advanced search strings with exact string matching, phrase matching, exclude words, “or”-operator etc.
  • The new search functionality is documented as search tips when the search field is focused on the Home page
  • Improved search results ranking

Improvements And Changes

• Individual task steps timeout after 30 mins to prevent endless running tasks
• Support for custom certificates when running Katalogue behind company proxies
• Database view v_job_step now show the individual job step runtime and start time

Bug Fixes

• Fixed: Datasource sync jobs get stuck on the “delete relationships” step and never complete
• Fixed: Testing the datasource connection when a custom port number is entered fails with an error related to port number datatype.
• Fixed: On some occasions, adding MSSQL datasources fails with an error related to username must be on a specific format when using a trusted connection, eventhough trusted connection is disabled.
• Fixed: Local users are logged out every 15 minutes with a “jwt expired” error message. This is due to the JWT access tokens are not refreshed correctly
• Fixed: Katalogue emails have a few formatting glitches in the Outlook Desktop app

v0.11.0

Released 2023-03-09

New Features

• Email notifications
  • Use either Azure or a custom SMTP server to let Katalogue send emails
  • Send system emails (currently only task run failures) to one or more email addresses
  • Email notifications are sent to users whenever their passwords are changed.
• Reset password functionality for local users. This lets users reset their passwords on their own. This feature requires email notifications to be enabled and configured.
• Send a welcome email to new users with an auto-generated password (for local users only) and login instructions. This feature requires email notifications to be enabled and configured.
• Improved Select inputs:
  • Select inputs to associate fields in Field and Field Description edit form has a “Select All” functionality
  • Improved “drop down” button design
• Improved Password inputs:
  • Caps Lock warning
  • Show/hide password button
• 404 not found page when following a link to an asset that does no longer exist

Improvements And Changes

• Better performance when loading Fields in the Dataset view
• SideDialog vertical scroll on content only. Title and Save-buttons remain static.
• Major code refactoring in frontend app to improve scalability, robustness and modularity
• A number of minor GUI changes and optimizations

Bug Fixes

• Fixed: The search index is not properly updated when a system is deleted, resulting in search results showing up that no longer exist
• Fixed: There is an error when trying to delete tasks
• Fixed: Datasource sync tasks sometimes fail with a timeout error. The timeout has been increased from 60 sec to 30 min
• Fixed: Local users are logged out every 15 minutes with a “jwt expired” error message. This is due to the JWT access tokens are not refreshed correctly
• Fixed: Some dialog forms do not reset correctly after an asset has been added
• Fixed: Some browsers were prevented to save user password on login screen
• Fixed: Tooltips do on some occasions not disappear correctly
• Fixed: Some icons do not render properly in Safari
• Fixed: Owned items counter in the tabs in User view do not summarize correctly
• Fixed: A datatype error sometimes occur when Port number is specified in a datasource connection
• Fixed: Select drop-down options are sorted in descending order (not ascending as intended) in some browsers

v0.10.0

Released 2022-09-29

New Features

• Changelog
  • First version of this changelog to show changes.
• Table Pagination. Tables with > 200 rows are now paginated for increased performance.
• Settings Page. Enabled GUI based management of some settings.
• Application Logging. The application now emits logs to log files and the output to console is standardized with log level and timestamps.

Improvements And Changes

• Added runtime to tasks overview

Bug Fixes

• Fixed: Bug that prevented map-rule relationships to be created.