> ## Documentation Index > Fetch the complete documentation index at: https://docs.dqlabs.ai/llms.txt > Use this file to discover all available pages before exploring further. # Remediation > How Prizm's Remediation settings configure exception writing, metadata capture, and external Spark compute for remediation workflows. ## Overview The **Remediation** section in Settings configures how Prizm handles exception writing, metadata capture, and compute resources for remediation workflows. It is organized into three sub-sections: Exception, Metadata, and Compute. Each sub-section has an enable/disable toggle. Settings fields appear only when the toggle is turned on. Navigate to **Settings → Remediation** to access this section. *** ## Exception The **Write Exception** toggle enables or disables exception writing for the organization. When enabled, the following configuration fields appear: | Field | Description | | ------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Sources** | The data sources Prizm writes exceptions for. Defaults to **All** — covers all connected sources. Use **Add Sources** to restrict to specific sources. | | **Row Limit** | Optional cap on the number of rows written per exception event. Leave blank for no limit. | | **Exception Categories** | The issue categories that trigger exception writes. Defaults to **all** categories. | Click **Save** after making changes. *** ## Metadata The **Metadata** toggle controls whether Prizm captures and stores metadata during remediation runs. Enable this setting to retain metadata context alongside exception records — useful for auditing and root-cause investigation. Click **Save** after toggling. *** ## Compute The **External Spark** toggle enables external Spark compute for remediation workflows. When enabled, the Schema field appears: | Field | Description | | ------------ | ------------------------------------------------------------------------------------------------------------------------ | | **Schema**\* | The schema name in your external Spark environment where Prizm writes intermediate results and exception data. Required. | Click **Save** after entering the schema name. Image The External Spark schema must exist in your Spark environment before configuring it here. Prizm does not create the schema automatically.