Why Am I Receiving a 'Did Not Receive Done or Success Response in Stream' Error with Llama3.2?

Below are some points to check and suggestions that may help resolve the issue:

Verify the Ollama Instance Connectivity

Although the error message ("Did not receive done or success response in stream") isn’t explicitly covered in our common issues, similar troubleshooting steps for connection and streaming problems apply. In many cases, intermittent stream issues can be traced to how n8n connects to the local Ollama instance. You might want to verify the following:

• Correct Base URL Configuration:
  Make sure that your Ollama credentials in n8n use an address that actually works in your Docker environment. For example, if n8n runs inside Docker and Ollama is running either on the host or in its own container, consider whether you should use http://localhost:11434, http://127.0.0.1:11434, or http://host.docker.internal:11434 (if on Linux, using the correct host mapping is also important).

• Docker Networking Setup:
  Since you’re running n8n in Docker on macOS, check that networking is configured so that n8n can reliably reach Ollama. If only n8n is containerized, use the --add-host host.docker.internal:host-gateway flag (or its equivalent if needed) when starting your container. See our documentation for Docker networking with n8n for guidance.

Consider Resource Constraints

Since your local machine is a MacBook Pro M2 with 8 GB of RAM, resource exhaustion or intermittent slowdowns might affect high-load operations such as managing large language model streams. Although this error may not directly point to insufficient RAM, keep an eye on system resource usage when the chain fails.

Examine the Streaming Behavior

The error indicates that the expected final signal (a “done” or “success” response) is not being received in the stream. This can mean that the Ollama server might be unexpectedly terminating the stream or that there is a timing/network issue:

• Server Stability:
  Ensure that the Ollama instance is stable and properly configured to stream responses via the endpoints expected by the node.

• Intermittent Response Handling:
  Because reruns sometimes work, the issue could be tied to a temporary hiccup in the communication between your Docker environment and your host. If possible, check logs on the Ollama side to see if there are any errors or warnings when the stream is interrupted.

Final Suggestions

1. Double-check your credentials and base URL: Confirm that you’re connecting to the correct address for your Docker setup. If you suspect IPv6 issues, switching from localhost to 127.0.0.1 can sometimes help.

2. Review network and container settings: Especially since Docker can create networking isolation, ensure that your containers are configured correctly and that any required Docker networking flags are in use.

3. Monitor system resources: On an 8 GB system, keep track of memory and CPU usage during these operations to rule out any resource-related interruptions.

4. Review Version Compatibility: Ensure that you’re using compatible versions of the n8n nodes and the Ollama model. Since you’re using “llama3.2”, confirm that the model version is fully supported by the node integration you’re using.

If after these checks the issue still persists, consider opening a support issue with detailed debug logs so that the n8n community or maintainers can further pinpoint whether it originates from the node integration code or the Ollama streaming service.