n8n local testing and explanation on Fedora
| Field | Value |
|---|---|
| Source | Obsidian · Claudian plugin |
| Session | ses_f60ed7bdeffeDFK6U7FTIHssG2 |
| Started | 2026-09-14 08:39 |
| Messages | 7 |
| Parts | 25 (text+patch+tool) · 5 reasoning · 4 tool calls |
| Model | opencode:big-pickle |
| Directory | /home/sambasene/obsidian/Vault |
Transcript
Section titled “Transcript”afterasterisk last time it’s time to test n8n locally, while you’re at it also write in this file explanation about n8n is, what it’s for, what it can do, what it can’t, why i would need it, what about when used locally, how to install it, i’m on fedora by the way
<linked_content path=“n8n local testing series.md” />
read — {"filePath": "/home/sambasene/obsidian/Vault/n8n local testing series.md"}
<path>/home/sambasene/obsidian/Vault/n8n local testing series.md</path><type>file</type><content>
(End of file - total 0 lines)</content>bash — date
Mon Sep 14 08:39:52 AM GMT 2026websearch — {"query": "n8n install Fedora Linux npm docker 2026"}
The file is empty. Let me write the full n8n introduction and local testing guide.
write — `{“filePath”: “/home/sambasene/obsidian/Vault/n8n local testing series.md”, “content”: “—\ncreated: 2026-09-14\ntags: [n8n, automation, self-hosted, workflow]\n—\n\n# n8n — Local Testing Series\n\n
…[truncated]`
Wrote file successfully.Done. The file [[n8n local testing series.md]] is now filled with everything you need to know before your first local test:
What’s covered:
- What n8n is — open-source, self-hosted workflow automation (Zapier/Make alternative you fully own)
- What it’s for — connecting APIs, automating tasks, AI orchestration, content pipelines, devops glue
- 400+ integrations — HTTP, Slack, Telegram, Postgres, OpenAI, Ollama, GitHub, etc.
- What it can’t do — not a full app framework, no real-time streaming, heavy compute isn’t its job, UI is editor-only
- Why you need it — fast prototyping, AI workflow composition, no SaaS tax, full data privacy
- Local vs server — local is for testing/prototyping, server is for 24/7 uptime
- 3 install options on Fedora:
sudo npm install n8n -g— simplest- Docker with volume mount — cleanest isolation
- Desktop app — most user-friendly
- First-time setup — owner account, first test workflow
- Systemd user service — auto-start on login
The series checklist at the bottom is ready for your next steps once you get it running.
after login into the n8n interface in the browser, it’s asking me for an ai provider.
n8n’s onboarding screen wants you to pick an AI provider so it can pre-configure the AI nodes. Here’s what to pick:
If you have Ollama running locally (free, no API key):
- Select Ollama
- Base URL:
http://localhost:11434 - Pick a model you have downloaded (like
llama3,gemma3, etc.)
This is the best option for local testing — no cost, no API keys, everything stays on your machine.
If you don’t have Ollama or want stronger models:
- Select OpenAI
- Paste an API key from platform.openai.com/api-keys
- Pick
gpt-4o-minifor cheap testing orgpt-4ofor better results
If you have an Anthropic key:
- Select Anthropic and paste your API key
Quick recommendation
Section titled “Quick recommendation”For local testing, Ollama is the move. You get:
- Zero cost per call
- Complete privacy
- Works offline
- No API key management
If you don’t have it installed yet:
curl -fsSL https://ollama.com/install.sh | shollama pull llama3.1:8bThen point n8n at http://localhost:11434.
Which option are you going with? And do you have Ollama already or need to set it up?
5 internal reasoning block(s) omitted — raw thinking is not exported (see archive index).