Immuta v2020.2.0 Release Notes
Immuta version 2020.2.0 was released June 30, 2020.
v2020.2.0 New Features
Databricks R and Scala support: Enables fine-grained (table, column-, row-, and cell-level) access controls across R and Scala in addition to Python and SQL in Databricks.
Flexible Databricks cluster configurations: Removes high concurrency cluster (HCC) and table ACL requirement.
Dynamic Enforcement in Snowflake: Policies can now be enabled natively in Snowflake without requiring queries to pass through the Immuta proxy. This will also provide the option for using secure views or regular views.
Native database/schema support: The Immuta Query Engine will support original schema/database names rather than all tables being in the
immutaschema. In Databricks, all data policies will be enforced transparently in Databricks against the original native
database.tablename rather than requiring it to live in the virtual
Schema Evolution: Enables Immuta to continuously monitor for changes and update tables registered in Immuta. This allows data engineering to occur (creating new tables and columns) without requiring manual catalog updates in Immuta.
WRITE workspaces in Databricks and EMR: Allows data analysts and scientists to collaborate in a Databricks or EMR/Hive database, managed via Immuta projects, for writing analytical results or to store feature engineering output without leaking data to other users. This feature also allows newly generated data in the sandbox to be re-shared through Immuta with policy inheritance.
Automated CCPA: A templated policy (like the HIPAA Safe Harbor policy) included in Immuta that enables organizations to comply with regulations by automating the manual steps across discovery of sensitive data, to apply proper de-identification techniques, and to demonstrate compliance using acknowledgement workflows.
Randomized Response (advanced privacy-enhancing technology):
- Protects from attacks to uncover sensitive information associated with an individual.
- Enables analysts to rapidly derive utility and value from the column, without worrying about re-identification.
Internal Sensitive Data Discovery auto-tags and classifies data completely internal to the software without reaching out to an external service.
Global Policy cloning allows users to clone existing Global Policies in order to edit them.
Custom certifications can be created against any Global Policy.
Descriptions can be added to purposes.
Support for creating per-user connection limits to the Immuta Query Engine.
Bulk adding users/groups to data sources is now possible.
Impersonation is now supported for the Immuta API.
Better resilience so that when data fingerprinting fails other downstream processes, like Sensitive Data Discovery, can still execute.
Better support for soft deleting
Discoveredtags and persisting that change across multiple Sensitive Data Detection executions.
Support for markdown in catalog discussions and column descriptions.
Better Query Engine support for impersonation to allow pre-SQL execution in dashboards such as Tableau.
- Policy import/export: allows migration of policies to/from Immuta instances and provides an integration point for source control around policies.
v2020.2.0 Major Bug Fixes
- Snowflake rounding policies will now use the specified bucket size.
- Snowflake Variant columns will no longer break fingerprint or high cardinality calculations.
- Fingerprint will no longer fail on out-of-range date values.
- Fix ambiguous column references on joins of data sources with K-Anonymization policies.
v2020.2.0 Known Bugs
v2020.2.0 Deprecation Notices
v2020.2.0 Breaking API Changes
/dataSource/jobswas removed from the API
v2020.2.0 Migration Notes
- Customers must be using at least Immuta v2.6 before migrating to v2020.2.0.
- Anyone using Native Snowflake in Immuta v2.8 will need to upgrade their Native Snowflake integration to the latest version on the App Settings page.
Immuta v2020.2 Patch Releases
See the following pages for details about each release: