Create Custom Monitors

Custom monitors let you define exactly which condition to watch for, on any metric available for your data source, and notify the right people when it's breached.

Navigating to Monitors

Navigate to Compute > Monitors. The Monitors list shows every monitor you've created or imported.

Hovering over a row also shows the monitor's help text (if provided) and how many notification channels it has configured.

Managing the Monitors list

With the Monitors list, you can:

• Sort – Select the sort icon on a column header to sort ascending or descending.

• Filter by field – Select the filter icon in the top-right corner of the table, choose a metric (Monitor Name or Updated On), select an expression, and enter a value.

• Filter by category – Use the left-pane filters to narrow the list by Entity, Data Source Name, Severity, or Status. Selecting multiple checkboxes applies an AND condition across them. Filters applied through the filter icon and the left pane combine together.

• Refresh – Select the refresh icon to load the most recent monitors.

• Activate or deactivate – Use the toggle in the Status column to turn a monitor on or off without deleting it.

• Edit – Select a monitor's name to open the Edit Monitor page, update its parameters, and select Update Monitor. If an edit doesn't appear on the list immediately, refresh the page.

• Delete – Open the monitor's Edit Monitor page and select Delete.

• Clone – Select the options button on a monitor row, choose Clone, then adjust the name and any settings before saving. The cloned monitor is a separate monitor referencing the original by name.

For monitors with a built-in recommendation, you can enable or disable the recommendation directly from Edit Monitor using its toggle. Recommendations expire automatically once they're no longer relevant.

Create a Monitor

A monitor evaluates a query against your data on a schedule you define. When the result crosses your configured threshold, ADOC raises an alert.

1. Navigate to Compute > Monitors.

2. Select Create Monitor.

3. Enter a Monitor Name. Names can't contain blank spaces.

4. Select a Data Source. If you have multiple instances of the same data source, you can target one instance or all of them.

5. Select an Entity — see Choosing an Entity below. The entities available depend on the data source you selected.

6. Define the Query:

   1. Under Aggr, select the arithmetic operation to apply to the metric, then select the metric from the dropdown. To evaluate more than one metric, add another aggregation.

   2. Under Filter by, narrow the data the monitor evaluates — for example, to a specific region or environment. You can add multiple filters.

   3. Under Group By, group results by a metric, such as cloud provider. Select Add Parent Aggregation to add a further grouping on top — for example, counting unique cloud providers in addition to grouping by them.

7. Configure the schedule, thresholds, and notification channels for the monitor.

Example: To monitor CPU usage limited to devices in a specific AWS region and grouped by cloud provider, set the aggregation to CPU usage, add a filter restricting the region, and group by cloud provider.

Choosing an entity

The entity defines what the monitor watches within the data source. Available entities depend on the data source you select.

Databricks entities

Snowflake entities

What's next

• Stock Monitors – Import predefined monitors instead of building one from scratch.

• Compute | Chargeback & Budget – Understand where monitor-triggered costs show up in your budgets.