# Case instances

> Real-time and historical view of Maestro Case executions, covering status tracking, activity volume, and failure investigation across all case instances in a tenant.

The **Case instances** page shows a real-time and historical view of all Maestro Case executions in your tenant. It helps you monitor activity volume, track status, and investigate running or failed case instances.

Use this view to:

* Identify which cases are used most often
* Spot failures or anomalies quickly
* Drill into instance-level execution data

The following diagram shows how Maestro displays case instances. It illustrates how each instance represents a single execution of a case and how users can monitor its status, duration, and key metrics in real time.

  ![case instances view](https://dev-assets.cms.uipath.com/assets/images/maestro/maestro-process-instances-view-588189-16f29462.webp)

## Key sections and fields

### Top cases by instances

A horizontal bar chart showing the cases with the most instance runs over the selected time range.

### Top cases by duration

Highlights cases with the longest average run times, helping you detect potentially inefficient case flows.

### Case instance status

A stacked bar chart showing the number of completed vs. faulted instances over time. Useful for identifying spikes in errors or system behavior trends.

### Case instance table

A sortable list of all cases with instance activity. Columns include:

* **Case name** – The name of the case.
* **Running** – Number of currently running instances.
* **Completed** – Number of instances that completed successfully.
* **Faulted** – Number of failed instances.
* **Total** – Sum of all instance states (running, completed, faulted).
* **Versions** – Number of case plan versions deployed.
* **Location** – Workspace or shared space where the case is published.

Use the **Search** and **Location** filters to narrow results to specific projects or teams.

:::note
For tenants with large numbers of case instances, Maestro displays initial records immediately while continuing to load remaining data in the background, so you can monitor, filter, and search without waiting for the full dataset. The progress indicator disappears once all available records are loaded, and your applied filters remain accurate as the complete dataset populates.
:::