Getting error while trying to use external MCP

[SandipCoder]

I have created a mcp connection from marketplace :

Conn Name : mcp_brightdata_web_surfer

mcp product url : https://brightdata.com/

I have attached the mcp through playground . But while submitting the prompt I am getting the following error .

Error : The endpoint Llama 4 Maverick has encountered an error and responded with the following:

{"error_code":"INTERNAL_ERROR","message":"{\"external_model_provider\":\"saas-llm\",\"external_model_error\":{\"error\":{\"type\":\"INTERNAL_ERROR\",\"message\":\"INTERNAL_ERROR: Failed to list tools from external MCP server dbc-45e6c2e8-ccf1.cloud.databricks.com/api/2.0/mcp/external/mcp_brightdata_web_surfer: INTERNAL_ERROR: External MCP server returned error status 502: 502 Bad Gateway\"}}}"}

 

[RahulPathakDBX]

Based on the error,  which is a "502 Bad Gateway" error indicates that a server sitting as a gateway or proxy received an invalid response from an upstream server. Here, Databricks is trying to communicate with the external MCP server associated with Bright Data, but the MCP server is either down or malfunctioning and cannot fulfill the request.

Possible Root Causes for the Databricks MCP/Bright Data Integration Error:

• Bright Data MCP or related APIs might be experiencing downtime or heavy traffic.
• There could be network issues between Databricks and Bright Data (e.g., firewall problems or expired SSL certificates).
• MCP connection settings may be incorrect, such as wrong credentials, endpoint URL, or missing permissions.
• Bright Data might be rate-limiting or throttling requests due to account quotas or resource limitations.
• Databricks Playground may not be fully compatible with Bright Data MCP, or there might be unannounced integration changes.
• Tool listing failures could result from schema mismatches, API changes, or misconfigured access on Bright Data’s side.

Initial Troubleshooting Steps:

• Check Bright Data’s network status page or contact their support to see if there are service outages.
• Verify your MCP connection configuration in Databricks; make sure product URL, credentials, and integration settings are correct.
• Test accessing the Bright Data MCP API directly (using tools like curl or Postman) to confirm connectivity.
• Review your Bright Data account’s quota and ensure you're not exceeding rate limits.
• Check for recent updates or changes from Databricks or Bright Data that may affect MCP integration.
• If these steps don't resolve the issue, provide error logs and configuration details to both Bright Data and Databricks support for further assistance.