V502 - V551

A new entry is added each week, and contains the release notes for all of the Tonic versions that were released during that week.

V541 - V551

August 12, 2022

Added a new Max Categorical Dimension parameter to the AI Synthesizer configuration. This parameter controls the dimension of each column that has categorical or location encoding. If a column contains more distinct categories than this parameter, the most frequent categories are embedded as distinct one-hot vectors. The remaining categories are combined into a single one-hot vector.

Improved error identification for an invalid where clause in subsetting configuration.

On the subsetting page, for tables that were not previously in the subset, the row count is now correctly represented as unknown instead of 0.

Fixed an issue with the Tonic update option in the Tonic application.

Configuring a subset target table to include 100% of the records no longer causes an error during data generation.

Removed connection pooling from Tonic workers to address database connection issues during data generation.


  • Added support for partial indexes.

  • Fixed an issue where the configured generators were not applied.

  • Indexes that use the collation option are now properly recreated in the destination database.

  • The subsetting user interface now uses the correct terminology for MongoDB.

  • Improved performance for Mongo subsetting when handling downstream tables.

  • Fixed percentage-based subsetting for Mongo versions before 4.4.2.


  • Added support for HASH partition parallelization.

  • Running data generation on masked and passthrough tables with ranged sub-partitions no longer results in duplicated data.

  • Added support for parallel uploads with ordering.


  • Added new environment variables (ORACLE_TRACE_LEVEL, ORACLE_TRACE_FILE_LOCATION, ORACLE_TRACE_FILE_MAX_SIZE, and ORACLE_TRACE_OPTION) to enable Oracle tracing.

  • Added support for datetime components in composite keys for subsetting.

  • Unique indexes are now detected and users cannot apply generators that might violate the enforced uniqueness.

  • Improved performance for data generation.

  • Improved handling of schema names during data generation.

  • Users are no longer incorrectly removed from the destination database.


  • Upgraded npgsql to address an issue with cross-schema types.


  • Added support for Kerberos authentication for HDFS with Spark / Livy.

  • Added support for repartition and coalesce options for Spark EMR and Livy.

  • Repartition and coalesce options can now be saved on Databricks.

  • Added support in Hive for varchar and char fields that have lengths.

V534 - V540

August 5, 2022


On Privacy Hub, the list of unprotected sensitive columns is replaced with a Database Tables list. The Database Tables list summarizes the protection status of each table in the source database. The Privacy Status column summarizes the protection status of the columns in the table. It provides access to the same column details and configuration options as the protection status panels at the top of Privacy Hub.

You can now filter the Workspaces view based on your assigned role in the workspace. For example, you can only display workspaces for which you are an Owner or an Editor. To use the role filter, click the filter icon in the Role column heading. For admin users, the role filter includes a Role assigned toggle to allow them to only see workspaces that they have a role in.

Other updates

Customers in the Basic license tier can now transfer ownership of workspaces and assign workspace roles to other users.

New users are now logged into Tonic immediately after they create their Tonic account.

For AI Synthesizer:

  • The modeling now incorporates static features for each entity across a sequence of events.

  • AI Synthesizer now queries to retrieve the actual minimum/maximum values of relevant columns, instead of taking the minimum and maximum of a sample of the data.

To update a license key, self-hosted instances that do not have an admin user can set the license key as the value of the TONIC_LICENSE environment variable. Tonic ignores the variable in instances that have an admin user.

The Workspaces view no longer waits for users to finish applying several new filters in quick succession before it fetches new results.

On the Job History view, fixed an issue where the copy job ID and download logs icons flashed on hover. Removed an error that flashed when job details were displayed.

Fixed the parallel processing for subsetting.


  • Indexes are now copied to the destination database.

  • Percentage-based subsetting is now supported in versions earlier than 4.4.2.

  • Deleted collections are no longer displayed.


  • Improved memory handling for uploaded CSV files.


  • Long and long raw columns are no longer converted to blob or clob.


  • Added Maven repository and artifact information to use the SDK Launch method to download the Spark SDK.

  • Improved the performance of data generation on Databricks when using Job Cluster. Added support for SQL Warehouses on Databricks.

V528 - V533

July 29, 2022

Added Sequence Length Loss Factor and Order Column Loss Factor model configuration options for events data to the AI Synthesizer. Sequence Length Loss Factor indicates the importance of realistic sequence lengths in the model. Order Column Loss Factor indicates the importance of realistic column value ordering in the model.

For the Categorical generator, differential privacy is now off by default.

Increased the amount of time after which an inactive job is assumed to be canceled.


  • Fixed a subsetting issue that caused errors when there was missing data.


  • Tonic now handles the BIT data type correctly.


  • Decimal values that are larger than the dotnet decimal data type can now be handled.


  • Fixed an issue that caused the system to crash when you clicked Test Connection during workspace creation.

  • Fixed an issue that prevented more than one generation run per version.

  • Fixed a HIPAA resource issue that caused data generation to fail.


  • Fixed an issue that caused the system to crash when you clicked Test Connection during workspace creation.

  • Fixed an issue that prevented more than one generation run per version.

  • Fixed a HIPAA resource issue that caused data generation to fail.


  • Fixed an issue with applying the Null generator to a struct array column.

SQL Server

  • Synonyms are now created correctly during data generation.

  • Fixed an issue with subsetting based on a percentage of a target table.

V520 - V527

July 22, 2022

New features and enhancements

Redesigned the user experience for the subsetting feature. The new subsetting view displays a list of the source database tables and indicates whether each table is in the subset. When you click a table, the table details display in a new right-hand panel. From the details panel, you can configure target tables. You can also identify lookup tables (previously referred to as reference tables), and indicate how to handle tables that are not in the subset. For details, see About subsetting.

If subsetting is configured, when you run a data generation job, you can enable or disable subsetting. For details, see Running a data generation job.

Added latitude and longitude processing to the HIPAA Address generator.

You can now filter Database View based on the applied generator. On the filter panel, in the Applied Generator field, you can provide the list of generators to include. See Assigned generator.

Redesigned the Schema Changes page. The Actions list is now called Conflicting Schema Issues. Tonic provides clearer warnings when a schema change resolution will result in a change to the workspace configuration. Resolving a conflicting issue or bulk dismissing non-conflicting issues now includes a confirmation step. See Viewing and resolving schema changes.

Other updates

Renamed the Events generator to Event Timestamps.

Fixed some small display issues in the new Privacy Hub and Subsetting displays.

Improved event generation for the AI Synthesizer.

Improved how we retrieve CloudWatch logs to include the job ID and to use the correct Tonic version.

Corrected the processing of downstream, multi-index tables during subsetting. Corrected an issue in the initial fix.

Improved memory usage during data generation.


  • Corrected how permissions are validated for Heroku.


  • Tonic now prevents data generation jobs from running when the Oracle versions for the source and destination database do not match.

  • Improved the performance of writes to the destination database.

  • After deleting a table in the destination database, Tonic now also clears out the recycle bin.

  • Fixed generation for LONG and LONG RAW fields.


  • Fixed test connection and generation in response to an AWS API change.


  • Corrected the handling of time zones in timestamps.

  • Fixed test connection and generation in response to an AWS API change.


  • UUIDs larger than 16 bytes are now truncated.

  • Improved the use of MongoDB resources.

  • Improved the display format of LUUIDs.

V509 - V519

July 15, 2022

New features and enhancements

We enhanced Privacy Hub to add expanded top-level metric panels. These panels show the number of unprotected sensitive columns, protected columns, and unprotected non-sensitive columns. From these panels, you can display column details, select and configure the column generator, view sample data, and add column comments. For more information, see Privacy Hub.

You can now add or update a Tonic license from the Tonic application. For a new instance, you are prompted to provide the license key before Tonic displays the login screen. Tonic displays a message when the current license is expired. The Admin Panel includes an option to update the license key. For more information, see Entering and updating your license key.

For the Address generator, added City with State and City with State Abbr to the available options for the column format. You use these options for column values such as San Francisco, California or Boston, MA.

Tonic now supports subsetting for MongoDB databases.

Workspaces can now have a description (up to 200 characters) as well as a name. Use the description field to provide additional context for the workspace and how it is used.

In the Tonic API, you can now sort workspaces based on the last generation date.

Other updates

Tonic now prevents a job from running when the worker determines that the server is running a different version of Tonic.

Refactored the underlying implementation of the subsetting feature.

Made some small memory improvements for data generation.

Implemented performance improvements when applying parallel constraints.

Corrected errors for edge cases related to the Audit Trail.

When the selected workspace changes, the identifier in the URL is now updated correctly.

The workspace configuration is migrated to a data type that enables more efficient querying.

Tonic now validates uploaded foreign keys against the table definitions.


  • A destination database can now have more than 1000 tables that have the table mode set to Preserve Destination.


  • Batch sizes are now set dynamically based on the average row size.

  • Memory improvements for PostgreSQL data generation that involve large rows.


  • Improved support for foreign keys.

  • Added support for Apache Livy and HDFS.

  • Improved performance and added SDK support for the Integer Key generator.


  • MongoDB aggregations can now use temporary files on disk to store data that exceeds the MongoDB size limit. This expands the possible generations for a MongoDB database.

  • Corrected the generation and display of UUIDs.

  • ObjectIds can now be used as primary keys for subsetting in MongoDB.

  • Improved the usability of the Add Foreign Key Relationships tab on the Foreign Key Relationships view.


  • Improved how Tonic handles maximum lengths when it generates the following data types:



July 5, 2022

SQL Server

  • Eliminated duplicate default constraint URNs during database creation.

V503 - V507

July 1, 2022

Refreshed the Audit Trail user interface on Privacy Hub. The new Protection Audit Trail provides a paginated list of the updates to the sensitivity designation and generator assignments.

Deep links now work correctly when you use Google SSO to authenticate.

Error messages from Oracle are now displayed in response to invalid where clauses in subset configuration.

Made minor memory improvements to the Address generator.


  • Reduced the frequency of schema change detection on Snowflake databases. This can result in cost savings on Snowflake clusters, because the clusters can sleep more often.

  • Snowflake generation now works correctly when there are foreign key constraints.

SQL Server

  • Added support for security policies, sequences, check constraints, and system versioned temporal tables.

V501 - V502

June 24, 2022

Improved cross-tab support for automatic logouts when you configure an inactivity period.

The Update option in the actions menu now takes you directly to the System tab on the Admin Panel instead of the Users tab.

Corrected the password length requirement to be 12 characters or greater instead of greater than 12 characters.

Improved the estimated row progress for scaled tables.

Eliminated a race condition that occurred when applying constraints.

SQL Server:

  • The Categorical generator can now support more than 2 billion rows in a category.


  • Can now run concurrent jobs that use different versions of Databricks.

  • Tonic now supports ORC and Hive tables in Databricks.

Last updated