v0.195.0
Summary
Release v0.195.0 of the Databricks Go SDK adds an `UndropMetastoreById` method to the workspace-level Metastores service, which restores a dropped metastore by its ID. The release notes list no other features, fixes, or breaking changes.
Summary generated by brickster.ai. For the full changelog and any code/binary attachments, follow the GitHub link above.
More from databricks/databricks-sdk-go
v0.194.0 adds a new lakewatch package with a workspace-level w.Lakewatch service, batch methods on w.AiFunctions (CreateBatch, GetBatch, ListBatches, CancelBatch), and feature spec methods on w.FeatureEngineering (CreateFeatureSpec, GetFeatureSpec, DeleteFeatureSpec). It is a breaking change that removes the Critical value from jobs.AiRuntimeTaskPriorityClass, so code that references that constant will fail to compile.
Release v0.193.0 of the Databricks Go SDK adds a ViewName field to ml.IngestionDestination and a Notifications field to ml.MaterializationSet. The release contains only these API additions, with no fixes or breaking changes.
v0.192.0 adds new ComputeMigrations and GenieAgents workspace services, Sandbox file upload and download methods, and job-completion triggers and serverless compute options for Jobs. It also breaks the iamv2 API: many request types drop their account, group, user, and principal ID fields in favor of Name and Parent fields, and the AccountIamV2 and WorkspaceIamV2 methods take arguments in a new order and have new paths.
Databricks CLI authentication no longer rejects small wrapper scripts, such as launcher scripts or version manager shims, as the legacy Python CLI. Small files in PATH are now rejected only if they are the pip-installed entry point for the legacy Python CLI, and a path set in DATABRICKS_CLI_PATH or databricks_cli_path skips this check.
Release v0.191.0 adds SDK methods for AgentKit memory and session store permissions, budget spend status queries, feature engineering materialization sets, and starting or stopping vector search endpoints. It also adds external spend inclusion for budgets, unhealthy-table notifications for data quality, relative error options for approximate count-distinct and percentile functions, and table trigger conditions and timing controls.