For Model Deployment Private Endpoints

Troubleshoot issues when creating or using private endpoints with model deployments.

Unable to Make an Inference Request

A timeout error during an inference request from the service or application might indicate a network connectivity issue.

Network connectivity is the issue.

To ensure the application can reach the private endpoint resource, validate whether private endpoint URL resolves to an IP address by running the following command from your instance or application:
nslookup <private-endpoint-url>

The Following Service Limits were Exceeded

When you try to create a private endpoint, you get the error:
The following service limits were exceeded: ds-private-endpoint-count. Request a service limit increase from the service limits page in the console

The private endpoint limit for the tenancy is reached.

Create a service limit increase request to increase your Data Science private endpoint limit. For more information, see the links in the Support card on the home page.

Private Endpoint Creation Failure

When you create a private endpoint, the workflow fails

The specified subnet doesn't have enough IP addresses. This situation can occur when the specified subnet is configured for a small CIDR block of IP addresses, and all IP addresses in that block are already used.

Resize the VCN and try again, or select a different subnet that has enough IP addresses. For more information, see Changing a VCN's IPv4 CIDR blocks and IP Address Insights