Integrating OllamaDeepSeek with n8n on a VM (Google Cloud)

This tutorial will guide you step-by-step to deploy Ollama on a virtual machine (VM) using Docker and connect it with n8n to automate workflows.

Requirements: Ensure your VM has at least 4 GB of RAM. If the VM does not have enough memory, the download of the deepseek-coder model will fail.

---

1. Prerequisites

• Docker installed on your VM.
• n8n deployed using Docker.
• Access to your VM's terminal.

Add Your User to the Docker Group

To run Docker commands without sudo, add your user to the Docker group:

sudo usermod -aG docker $USER

Note: After adding the user to the Docker group, log out and log back in for the changes to take effect.

Set Permissions for the n8n Volume

Ensure the correct permissions are set for the n8n data volume:

sudo chown -R 1000:1000 /var/lib/docker/volumes/n8n_data/_data
sudo chmod -R 700 /var/lib/docker/volumes/n8n_data/_data

This grants the appropriate permissions for n8n to access its data.

---

2. Download the Ollama Image

Run the following command to download the official Ollama image from Docker Hub.

docker pull ollama/ollama

This command downloads the latest version of Ollama.

---

3. Run the Ollama Container

Now run the Ollama container and expose port 11434.

docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama

Explanation:

• -d: Runs the container in detached mode.
• -v ollama:/root/.ollama: Creates a persistent volume for the downloaded models.
• -p 11434:11434: Maps port 11434 of the container to the VM.
• --name ollama: Names the container ollama.

---

4. Create a Docker Network for Container Communication

Create a custom Docker network named n8n_network.

docker network create n8n_network

This allows n8n and Ollama to communicate with each other.

---

5. Connect the Containers to the Custom Network

Connect both n8n and Ollama to the n8n_network.

Connect n8n to the network:

docker network connect n8n_network n8n

Connect Ollama to the network:

docker network connect n8n_network ollama

---

6. Verify the Connection Between Containers

Inspect the network to ensure both containers are connected.

docker network inspect n8n_network

You should see n8n and ollama listed in the containers section.

---

7. Manage Models in Ollama

a. Access the Ollama Container

docker exec -it ollama bash

b. List Downloaded Models

Inside the Ollama container, run:

ollama list

This will display the models you have downloaded.

c. Download and Run the deepseek-coder Model

If the model is not downloaded, run it to start the download:

ollama run deepseek-coder

d. Remove the deepseek-coder Model

If you need to remove the deepseek-coder model to free up space or update it, run the following command inside the Ollama container:

ollama rm deepseek-coder

e. Download a Specific Version of Ollama Model (e.g., version 3.2)

To get a specific version of an Ollama model, replace MODEL_NAME with the desired model name:

ollama run MODEL_NAME:3.2

For example, to download version 3.2 of a model named ollama, run:

ollama run llama3.2:1b

ollama run deepseek-r1:1.5b

f. Exit the Container

exit

e. Verify the Status of the Containers

To check if Ollama is running:

docker ps -a

If Ollama is stopped, restart it:

docker start ollama

If you want restart n8n stopped, restart it:

docker stop n8n

docker start n8n

---

8. Configure n8n to Connect to Ollama

a. Change the Base URL in the n8n Ollama Node

When configuring the Ollama node in n8n, ensure the base URL points to the container name Ollama instead of localhost.

Correct URL:

http://ollama:11434

b. Verify the Connection from n8n

To ensure n8n can communicate with Ollama:

1. Access the n8n container:

docker exec -it n8n bash

2. Run a curl command to test the connection:

curl http://ollama:11434

If you receive a response, the connection is working correctly.

---

9. Troubleshooting

• Issue: The Ollama container stops automatically.

  • Solution: Check the logs to identify the error:

    docker logs ollama

  • Ensure you have at least 4 GB of RAM on the VM.

• Issue: n8n cannot connect to Ollama.

  • Solution: Ensure both containers are on the n8n_network and use the URL http://ollama:11434.

---

Done! You now have Ollama running locally on your VM and connected with n8n to automate your workflows.