• Ready for review
  • Workspace Cost Overview guide

    Introduction

    The Workspace Cost Overview gives the user the bird's eye view of everything in their subscription. This view is often where a user will start when looking for issues to investigate or features to improve.

    Reading left to right and top to bottom each widget gets progressively deeper as far as what data it is showing the user - the upper left shows overall cost and budget, and the lower left shows exactly what that cost breaks down into.

    Subscription Totals

    In the top left the user can see their overall totals at a glance to serve as a foundation for the graphs found below:

    • MoM Change: Total change in cost between the current month and the month before - shown in both dollar amount and percentage change.

    • Current AVG / Day: Average cost per day since the start of the fiscal year.

    • Current Month Total: Total cost of the lakehouse infrastructure for the current month.

    • Cloud Monthly Total: Total cost of the cloud infrastructure (e.g. VMs, storage, network) for the current month.

    • Databricks Monthly Total: Total cost of the Databricks infrastructure (e.g. workflows, clusters, SQL warehouses) for the current month.

    Subscription Cost

    The subscription cost graph focuses on the total amount spent and whether that amount fell within the set budget and/or cloud commit set.

    Monthly View

    In the monthly view, the user can see the total subscription cost of each month over the twelve-month period of the selected time range (last 12 months, current year, or fiscal year).

    If a month was within the set budget is shown in green, and if was over budget that month was shown in brown. The budget itself is shown as a yellow line on the graph - in this example, it is set to $500.

    Setting the Budget and Cloud Commit

    Both the Budget and Cloud Commit can be set by clicking the gear icon in the upper left corner of the widget. This will open a dialog where the user can set the annual budget and cloud commit values, along with a custom Fiscal Start Date for organizations that might have an unorthodox starting point for their fiscal year.

    Cumulative View

    Switching to the cumulative view will give the same information as the monthly view, but each month will show the cumulative amount spent for the year.

    This allows the user to see the overall cost month to month compared to the actual annual budget. This is helpful because the monthly budget is only the annual budget divided by 12, but some months might be more expensive than others. If all that really matters is that the annual total comes in under budget, this view might be more informative.

    This graph also has the added perk of showing a cost trend line to help forecast if the cost will come in under budget for the fiscal year - in the above example, despite some months being “over budget” the fiscal year is actually going to come in under budget ending the fiscal year at the end of September.

    Azure vs Databricks Cost

    The Azure vs Databricks breaks down the monthly cost by exactly what it says: Azure costs and Databricks costs - how much is the customer spending on each service?

    Hovering over a specific month will give the user exact dollar values of the individual costs for Azure and Databricks respectively, along with the percentage that amount makes up of the customer's total subscription cost.

    Further to the right, the user can also see the month-to-month cost change in dollar amount and percentage as well - highlighted in green if it went down since last month, or red if it went up since last month.

    Cost Breakdown

    The Cost Breakdown widget shows the user a total breakdown of the selected time period (current month by default) by what services/resources or features the subscription cost breaks down into - in other words: What specific things is the user paying for from Azure or Databricks?

    Subscription View (Azure) or Account View (AWS)

    Here the user can see exactly what resources and features make up the cost of their total subscription. What is being spent on VMs? Is the customer spending more on Job Clusters or All Purpose Compute? Is any money being wasted on Idle VMs?

    Workspace View

    If the user would prefer to view a list by workspace they can also do that here and see a breakdown of how much of the total cost of each workspace breaks down into Azure or Databricks cost.

    Workspace Stats

    The Workspace Stats chart takes all the information in the Cost Breakdown chart and visualizes it over time.

    Feature View

    Starting with the feature view the user can see all entities listed in the Cost Breakdown Feature list applied as a stacked bar chart.

    Hovering over any given month will once again give you a full breakdown of each element:

    • Amount in Dollars - Cost/amount spent

    • Amount in Percent - Percent of total monthly subscription cost for that feature

    • MoM Change in Dollars - Difference in cost from last month in dollar amount

    • MoM Change in Percent - Difference in cost from last month as a percentage

    Resource View

    The resource view will give the monthly breakdown by resources. The user can filter by the resouces they want to see.

    Hovering over any given month will once again give you a full breakdown of each element:

    • Amount in Dollars - Cost/amount spent

    • Amount in Percent - Percent of total monthly subscription cost for that feature

    • MoM Change in Dollars - Difference in cost from last month in dollar amount

    • MoM Change in Percent - Difference in cost from last month as a percentage

    Case Study: How much am I spending on Idle VMs?

    If a user wanted to investigate how much was being wasted on Idle VMs then a good place to start would be the Workplace Stats widget. By hovering over any given month they can see any cost that went toward Idle VMs in the listing, and if that increased since last month.

    Once they know a certain month incurred Idle VM costs, they can jump over to the Pools Report page and investigate the exact cause. For more on this topic, see the Pool Reporting docs.