Skip to main content

Release notes 26.3.4

ยท 2 min read

Dataedo 26.3.4 focuses on resolving performance issues and common bugs.

Portal:

  • Classification:
    • data-based Classification now runs faster on tables with fewer rows than the configured number of sample rows. Their columns are classified using the first sample query rather than an additional query per column
    • data-based Classification now skips objects that are broken or inaccessible, with a single readable warning. This applies to Snowflake, SQL Server, Azure Synapse, PostgreSQL, Amazon Redshift, MySQL, MariaDB and Databricks

Connectors:

  • Snowflake Bugfixes:
    • dynamic Iceberg tables are now imported as views with the Dynamic table subtype, including their definitions, columns, descriptions and lineage
    • the SQL parser now supports CREATE DYNAMIC ICEBERG TABLE definitions, so they no longer generate syntax errors in the parsing log
    • Data Profiling and Data Quality errors for broken or inaccessible objects are now displayed as readable messages instead of raw driver output

API:

  • the API endpoints for link metadata (/api/text-editor/link) and lossy description detection (/api/text-editor/detect-lossy-description) now require authentication

Agent:

  • when repository validation fails at startup, the Agent now logs the reason before shutting down
  • the Agent now executes only scheduled jobs created by Dataedo. Any other job found in the repository's scheduler tables is rejected

Desktop:

  • Custom Fields: Desktop no longer crashes when it displays a Tags custom field with multi-word tags that share a word (e.g., Asset Register, Asset). Each saved tag is now presented as a single token, with spaces replaced by _ and duplicates hidden