Connector Lifecycle
What happens if I delete a connector?
What happens if I delete a connector?
Deleting a source triggers an approval workflow based on your organization’s settings. Once approved, the source and all associated metadata are permanently deleted. The deletion cannot be undone after approval.
What happens to existing data and alerts if a connector is paused or disconnected?
What happens to existing data and alerts if a connector is paused or disconnected?
Existing metadata, quality scores, and alerts are retained while the connector is inactive. No new data is collected until the connector is re-enabled.
How long does Prizm retain historical data?
How long does Prizm retain historical data?
The default data retention window for time-series observability data and quality score history is 90 days.
Setup & Configuration
My cluster does not appear in the dropdown.
My cluster does not appear in the dropdown.
Only clusters that are currently running or resizing are listed. Start your cluster, then refresh the dropdown.
No containers appear when I configure my source.
No containers appear when I configure my source.
Your ADLS credentials are missing list permission on the storage account. Confirm the identity has Storage Blob Data Reader (source container) or Storage Blob Data Contributor (managed-storage container) as appropriate.
I selected a container and it disappeared from my selection.
I selected a container and it disappeared from my selection.
Your credentials don’t have read access to that specific container. The dropdown lists every container in the account because Prizm can’t filter it to read-access-only containers for a service principal or access token — but Prizm still checks read access on selection. If it fails, a toast appears and the container is auto-removed. Choose a container your credentials can read, or ask your Azure administrator to grant access.
Can I use Amazon EMR or Cloudera instead of Databricks?
Can I use Amazon EMR or Cloudera instead of Databricks?
Not at this time. Databricks is the only supported external Spark engine today.
What if my Databricks environment cannot reach ADLS?
What if my Databricks environment cannot reach ADLS?
This is handled through agent-outbound connectivity rather than manual firewall rules. Confirm agent-outbound is correctly configured and reachable for your environment. If jobs still fail to reach ADLS, contact your Prizm account team.
Authentication & Permissions
My jobs started failing after working correctly for weeks.
My jobs started failing after working correctly for weeks.
If you’re using a Service Principal, check whether the Azure client secret has expired — an expired secret prevents Prizm from reading ADLS. If you’re using Managed Identity, there’s no secret to expire — check instead whether the role assignments on the Access Connector are still in place.
Reads work but writes to the managed-storage container fail.
Reads work but writes to the managed-storage container fail.
This almost always means the identity has Storage Blob Data Reader instead of Storage Blob Data Contributor on the write container, or is missing Storage Blob Delegator at the storage-account level. Confirm both are assigned.
Credential vending fails, or Unity Catalog cannot issue a SAS token for the external location.
Credential vending fails, or Unity Catalog cannot issue a SAS token for the external location.
Storage Blob Delegator must be assigned at the storage-account scope, not the container scope — a container-scoped assignment will not enable delegation-key issuance. Re-check the scope of the role assignment, not just its presence.
Trino cannot see the Iceberg table.
Trino cannot see the Iceberg table.
Confirm external data access is enabled at the metastore level and the Trino principal has
EXTERNAL USE SCHEMA on the schema — table-level grants alone are not sufficient for external engine access.Troubleshooting
For connection errors, permission issues, and performance problems.
Glossary
For definitions of terms used in this FAQ.