chore(deps): update databricks-sdk requirement from >=0.109.0 to >=0.117.0

[dependabot]

Updates the requirements on databricks-sdk to permit the latest version.

Changelog

Sourced from databricks-sdk's changelog.

Release v0.117.0 (2026-06-11)

API Changes

• Add type_overrides field for databricks.sdk.service.database.SyncedTableSpec.
• Add type_overrides field for databricks.sdk.service.postgres.SyncedTableSyncedTableSpec.
• Change resource_id field for databricks.sdk.service.bundledeployments.Operation to no longer be required.
• [Breaking] Change resource_id field for databricks.sdk.service.bundledeployments.Operation to no longer be required.

Bug Fixes

• Cache tokens minted by DatabricksOidcTokenSource (Workload Identity Federation / account-wide token federation). Previously a fresh /oidc/v1/token exchange was performed on every authenticated API call, adding latency, amplifying transient federation-policy errors, and hitting OIDC token-endpoint rate limits. The token source now reuses the cached token until it is stale or expired, fetching a fresh ID token on each refresh to support rotation.
• Make WorkspaceClient.dbutils lazy (functools.cached_property). The constructor no longer imports databricks.sdk.runtime eagerly, so on Spark Connect (shared-access-mode) clusters, consumers that never read .dbutils — such as dbt-databricks Python models — no longer crash with CONTEXT_UNAVAILABLE_FOR_REMOTE_CLIENT during client construction.
• Fall back to the remote runtime implementation when the legacy user namespace cannot be materialized. On Spark Connect runtimes (e.g. shared-access-mode clusters), importing databricks.sdk.runtime — which happens when constructing a WorkspaceClient on such a cluster — tried to build a legacy SparkContext and raised CONTEXT_UNAVAILABLE_FOR_REMOTE_CLIENT at import time. It now logs a warning and falls back to the Spark Connect-compatible remote implementation instead of crashing.

Internal Changes

• Declare urllib3 as an explicit dependency. The SDK imports it directly; previously it was only available transitively through requests. No resolution change for users.

Release v0.116.0 (2026-06-10)

API Changes

• Add databricks.sdk.service.aisearch and databricks.sdk.service.bundledeployments packages.
• Add w.ai_search workspace-level service.
• Add w.bundle_deployments workspace-level service.
• Add running_instances field for databricks.sdk.service.apps.ApplicationStatus.
• Add custom_max_retention_hours field for databricks.sdk.service.catalog.CatalogInfo.
• Add environment_settings field for databricks.sdk.service.catalog.ConnectionInfo.
• Add custom_max_retention_hours field for databricks.sdk.service.catalog.CreateCatalog.
• Add environment_settings field for databricks.sdk.service.catalog.CreateConnection.
• Add custom_max_retention_hours field for databricks.sdk.service.catalog.CreateSchema.
• Add custom_max_retention_hours field for databricks.sdk.service.catalog.SchemaInfo.
• Add custom_max_retention_hours field for databricks.sdk.service.catalog.UpdateCatalog.

... (truncated)

Commits

• 8de8d38 Release databricks-sdk-py v0.117.0 (#1472)
• 8d80062 Fall back to remote runtime on Spark Connect when the legacy namespace is una...
• d01f89a Make WorkspaceClient.dbutils lazy via cached_property (#1470)
• c05cc6f Cache tokens minted by DatabricksOidcTokenSource (#1460)
• 528878e [Release] Release v0.116.0
• 1399208 Update SDK API to 55d22d7e7879e6be815cfe8e3e4cf7130f822c94 (#1471)
• 9744f91 Update SDK API to 805d88b51faa37301c51d1848da2db5a2e6ed921 (#1466)
• c897178 [Release] Release v0.115.0
• fef9b73 Fix error tracker URL to point to the Python SDK repo (#1462)
• d18e039 Honor AI_AGENT and pass raw values through (#1454)
• Additional commits viewable in compare view