# Managing Cache in the SQL Storage of the Reporting REST Service

## Environment

<table>
    <tbody>
        <tr>
            <td> Product </td>
            <td>
                Reporting
            </td>
        </tr>
    </tbody>
</table>

## Description

I encountered an issue where the cache in Telerik Reporting is not clearing automatically, causing certain database tables(e.g. **tr_String**, **tr_Object**) to grow indefinitely.

Each report generated is unique, so caching doesn't provide much benefit in this setup. I need to either prevent caching entirely or ensure the cache is cleared to prevent database growth.

## Solution

The Reporting REST Service will *always* periodically trigger cache clean-ups. However, depending on how the service is used, the conditions required for a given asset to be considered **expired** in the cache may need more time to be met. We advise looking into the [Telerik Reporting REST Service Storage Revealed](https://www.telerik.com/blogs/telerik-reporting-rest-service-storage-revealed) blog post for more details on the topic.

With that being said, in the sections below, we will share some of the recommended approaches for clearing the storage more often.

### Adjusting the `ClientSessionTimeout` Property

Use the [ClientSessionTimeout](https://www.telerik.com/products/reporting/documentation/api/telerik.reporting.services.reportserviceconfiguration#telerik_reporting_services_reportserviceconfiguration_clientsessiontimeout) property to control the session expiration time. The default value is **15 minutes**.

* Avoid setting it to zero, as this causes sessions to expire immediately, leading to errors like "410 Gone" when rendering reports.
* Use a reasonable timeout value to balance performance and user experience.

### Manual Cache Cleanup

To manually clear cache data, use SQL queries during maintenance windows or off-peak hours.

For example, execute the following query to identify and remove cache records based on a specific date threshold from the `tr_Oject` and `tr_String` tables. Start with a `SELECT` query to verify the results before running a `DELETE`.

#### SQL Query Example

```sql
SELECT TOP (1000) o.[Id], o.[Value]
FROM [RESTDB].[dbo].[tr_Object] o
INNER JOIN [RESTDB].[dbo].[tr_String] s
    ON LEFT(o.[Id], CHARINDEX('\', o.[Id], CHARINDEX('\', o.[Id]) + 1) - 1)
     = LEFT(s.[Id], CHARINDEX('\', s.[Id], CHARINDEX('\', s.[Id]) + 1) - 1)
WHERE s.[Id] LIKE '%\%\LCT'
  AND s.[Value] < '2026-01-01';
```

* Replace `'2026-01-01'` with an appropriate threshold date.
* After verifying the results, replace `SELECT` with `DELETE` to remove the records.

#### Additional Considerations

* Run manual cleanup during scheduled maintenance or off-peak hours to minimize disruptions.
* Ensure the `Reporting REST Service` is active for automatic cache cleanup to function correctly.

## See Also

* [Report Viewer and Reporting REST Services Cache Management](https://www.telerik.com/products/reporting/documentation/embedding-reports/html5-viewer-telerik-rest-service)
* [Telerik Reporting REST Service Storage Revealed](https://www.telerik.com/blogs/telerik-reporting-rest-service-storage-revealed)
