Running n8n workflow automation tool in docker-compose with OrbStack
Optically install n8n cli
# install CLI (one-time)
pnpm add -g n8n@next
# to update to latest version
pnpm update -g n8n@next To start n8n with PostgreSQL simply start docker-compose by executing the following
command in the current folder.
Important
But before you do, copy .env.example as .env in the project root directory,
then change the default postgres username and password in the .env file, if you prefer.
also check each n8n credentials files, and replace FIXME with appropriate values.
(optional) If you are using OrbStack, set this environment variables first:
export COMPOSE_PROJECT_NAME=agentic
export COMPOSE_FILE=compose.orbstack.yml
export COMPOSE_REMOVE_ORPHANS=true# for first time
docker compose --profile first-time-only up
# for next time
docker compose up
# if you want to ssh to n8n container, and copy files for RAG
docker compose exec -it n8n sh
mv /data/shared/my-resume.pdf /data/shared/bank_statementsAccess n8n dashboard at:
open https://n8n.agentic.orb.localAccess qdrant dashboard at:
open https://qdrant.agentic.orb.local/dashboard#/welcomeNote
First time when you access, you will see: Set up owner account page.
Enter some fake data, then you will be redirected to Dashboard
docker compose down
# DANGER: if you want stop and also want to remove all data, to reset database:
docker compose down -vThe default name of the database, user and password for PostgreSQL can be changed in the .env file in the current directory.
If you're running Ollama locally on your Mac (not in Docker), make sure OLLAMA_HOST environment variable set to host.docker.internal:11434 in the n8n service configuration in compose.yaml .
Additionally, verify ollama base URL is set to http://host.docker.internal:11434 :
Head to https://n8n.agentic.orb.local/home/credentials
Click on "Local Ollama service"
Change the base URL to http://host.docker.internal:11434
Note
If you are using LM Studio, use OpenAI Model integration and use lms url: http://host.docker.internal:1234/v1 URL instead of ollama url: http://host.docker.internal:11434
Important
If you are using OpenAi models, also verify OpenAi account credentials are set properly.
Quick start and usage
- Open http://n8n.agentic.orb.local in your browser to set up n8n. You’ll only have to do this once.
- Open the included workflow: http://n8n.agentic.orb.loca/workflow/srOnR8PAY3u4RSwb
- Select Test workflow to start running the workflow.
- If this is the first time you’re running the workflow, you may need to wait until
Ollamafinishes downloadingLlama3.2.
use n8n-cli to manage n8n server metadata
default location of n8n CLI config file: ~/.n8n/config
# First export env. NOTE: adjust db username/password as needed
export DB_TYPE=postgresdb DB_POSTGRESDB_HOST=localhost DB_POSTGRESDB_PORT=5432 DB_POSTGRESDB_DATABASE=n8n DB_POSTGRESDB_USER=user_postgres DB_POSTGRESDB_PASSWORD=user_passwordNote
If you haven't installed n8n cli, use docker compose exec -T n8n n8n in place of n8n for following commands.
in this case, output will be in container directories
# Export all the credentials to a specific directory using the --backup flag (details above):
n8n export:credentials --backup --output=config/n8n/credentials/
# Export all the credentials in plain text format.
n8n export:credentials --backup --decrypted --output=config/n8n/credentials/
# Import all the credentials files as JSON from the specified directory:
n8n import:credentials --separate --input=config/n8n/credentials/# Export all your workflows to the standard output (terminal):
n8n export:workflow --all
# Export all the workflows to a specific directory using the --backup flag (details above):
n8n export:workflow --backup --output=config/n8n/workflows/
# Import all the workflow files as JSON from the specified directory:
n8n import:workflow --separate --input=config/n8n/workflows/n8n audit- n8n Community videos and channels
- Self-hosted AI starter kit, video
- Deploy n8n with Docker Compose for Automating AI Workflows
- Install n8n The Right Way for You
- Local AI with Docker, n8n, Qdrant, and Ollama - Video
- Extract personal data with self-hosted LLM Mistral NeMo
- Building RAG Chatbot for Movie Recommendations with Qdrant and Open AI
- Build a custom knowledge RAG chatbot using n8n
- Scaling n8n, video
- Auth with n8n webhooks
- How I Parse 99% of PDFs into Structured Data