DLT DatePlane Error

[Stellar]

Hi everyone,

I am trying to build the pipeline but when I run it I receive an error

DataPlaneException: Failed to start the DLT service on the cluster. Please check the driver logs for more details or contact Databricks support.

This is from the driver log: ERROR Uncaught throwable from user code: com.databricks.sql.managedcatalog.UnityCatalogServiceException: [RequestId=d4068d65-d6c3-4070-b056-31cfdcaeb40f ErrorClass=INVALID_STATE.UC_CLOUD_STORAGE_ACCESS_FAILURE] Failed to access cloud storage: AbfsRestOperationException exceptionTraceId=bb83ca0d-09e5-48af-9c96-20e4acb63fa8
at com.databricks.managedcatalog.ErrorDetailsHandler.wrapServiceException(ErrorDetailsHandler.scala:33)
at com.databricks.managedcatalog.ErrorDetailsHandler.wrapServiceException$(ErrorDetailsHandler.scala:23)
at com.databricks.managedcatalog.ManagedCatalogClientImpl.wrapServiceException(ManagedCatalogClientImpl.scala:106)

What does this mean? I am trying with autloader to read the file with cdc logs and also this external location is set up and working properly.

Thanks!

[Kaniz]

Hi @Stellar, The error message you’re encountering indicates an issue with starting the Delta Live Tables (DLT) service on your cluster.

Let’s break it down:

1. DataPlaneException: This is a generic exception related to data plane operations.
2. Failed to start the DLT service on the cluster: The DLT service, which is responsible for managing data pipelines, failed to initialize.
3. UC_CLOUD_STORAGE_ACCESS_FAILURE: This specific error suggests a problem related to accessing cloud storage.

Here are some steps to troubleshoot and resolve the issue:

1. Check Cluster Configuration:

   • Ensure that your cluster configuration is correct. Verify that the cluster has the necessary resources (CPU, memory, etc.) to run the DLT service.
   • Confirm that the cluster is using the appropriate virtual machine type (VM type) and that it’s not exceeding any quotas (e.g., vCPUs).

2. Driver Logs:

   • The driver logs contain additional details about the error. You mentioned that you’ve checked the driver logs, but let’s dive deeper into them.
   • Look for any specific error messages related to cloud storage access or DLT initialization.

3. Cloud Storage Access:

   • The error message indicates an issue with accessing cloud storage. Make sure that the external location (where your CDC logs are stored) is correctly configured.
   • Verify the credentials (such as access keys or tokens) required to access the storage account.
   • Double-check the path to the CDC logs in the external location.

4. Quotas and Resource Limits:

   • As mentioned earlier, ensure that your cluster is not hitting any resource quotas (e.g., vCPUs, memory, storage).
   • If you’re using Azure Databricks, check your subscription type. Free trial subscriptions have limitations on the number of cores available.

5. Explicit vCPU Type (Azure Databricks):

   • If you’re using Azure Databricks, consider adding an explicit vCPU type to your cluster settings.
   • Edit the settings.json file and specify a valid vCPU type (e.g., "node_type_id": "Standard_DS3_v2").
   • Note that this change might require modifying the JSON version of the settings, as the UI version ma...¹.