2.22.0 Release Notes

✨ Release Highlights

• Unity Catalog Migration Assessment extended to unpublished workspaces

• Spend Insights / Workspace cost breakdown

• Spend Insights / Planning

• Spend Insights / Workspace Cost By Month

Open

Spend Insights main menu

📊 Spend Insights / Workspace cost breakdown

• cost breakdown by workspace and by feature definition

• find cost outliers in a workspace by feature categories

  • e.g. find the workspace with the most usage of All-Purpose Compute in the past 3 months

Open

Workspace cost breakdown

• support cost breakdown details (cost breakdown toggle) to display Cloud vs Databricks cost per feature category

• cost distribution by resource per workspace

• cost distribution by provider (cloud vs Databricks) per workspace

🏦 Spend Insights / Planning

• view spend by Year to Date or for previous years

• configure budget per month

• configure yearly commit for Cloud or Databricks

• compare actual and projected spend for current month

• view projected spend for next month

• preview top spend by tag group

• compare top spend by month or by week for complete periods

• top workspaces by spend

📊 Spend Insights / Cost By Month

• cost by month trends with detailed cost breakdown

• supports MoM $ and % trends

• customizable graphs to enable cost analysis

  • enable / disable rows to show on the bar chart

• support Nat gateway and Storage breakdown, currently with selection disabled

• cost breakdown by resource

🔭 Unity Catalog Assessment

• 11743 - ucm unpublished workspace

❤️ Incidents // Cost

• 11582 - sql warehouse cost alert

❤️ Incidents // Performance

• 11610 - Update root cause message format for cluster idle time incidents to include cumulated idleness and initial state of the problem

  • e.g. The cluster instance was idle for more than 8 days 2 hours 49 minutes, which exceeds the threshold of 1 hour. The problem was first detected after 1 hour 17 minutes of idle time.

📊 Analysis Optimizations

• 11659 - Retrieve cluster configuration from the APC configuration timeline instead of the Dbx API when analyzing cluster instances

• 11703 - Add logging for unresolved cost records (workspaceId or unidentified resource type)

🪩 UX Optimizations

• 11643 - on the consumption history page set the page size by default to 100

• 11208 - navigable (clickable) path navigation for Workloads

• 11684 - cost and performance tabs on incidents page are always expanded

• 11136 - included a dropdown in incident policies to navigate directly to other entities

• 11522 - Menu Settings // Unity Catalog Migration, Analysis Management, and Provisioning

🔔 Notifications

• 11623 - last system toast is saved in notifications

• 11810 - latest performance and cost incidents notification

Performance Optimizations

• 11812 - added indexes for speeding up the daily cost aggregation

🎛️ Configuration & Provisioning

• 11608 - Spark conf should contains only secrets and init script. All other spark-conf values moved to init script in order to not pollute user configuration

• 11660 - AutoPilot provisioning logs display both enabled/disabled items

• 11716 - Use Managed identity for keyvault code in provisioning flow

• 11757 - Expose version endpoint to not require authentication.

📈 Operational Management // Consumption (Cost) Analysis

• 11723 - Consumption history view update

  • collapsable sections for steps that process consumption data

🐞 Fixed Bugs

• 11649 - APC show notebooks/jobs navigation has wrong dates and appIds filter

• 11727 - Cluster instance - Workers number not displayed