Update deps and add Ollama model (#20)

* fix deps
* add Ollama service
* use plan id from public cloud
* remove tunnel and use internal route

Author: clemensfreitagdev

README.md

@@ -1,10 +1,21 @@
 # Instruction for deployment on Codesphere
 
-1) create a new workspace
-   insert this github link to clone this repo directly in your workspace: https://github.com/Datata1/deploy.n8n.codesphere
+1. Create a new workspace: 
+- Use the HTTP url of this repo to create a New Workspace in Codesphere
 
-2) go to the CI-Pipeline UI:
-   run the prepare-stage (this might take a couple minutes)
-   when the preparestage is ready (when it is green) run the run stage
+2. Go to "CI & Deploy": 
+- Run the Prepare Stage (this might take a couple minutes)
+- When the Prepare Stage is ready, start the Run Stage
 
-3) in the top right corner you can see 'open deployment'. cliuck this button and you can log in to n8n.
+3. Use 'Open Deployment' in the top right corner to open N8N.
+
+## Bonus: Ollama Model hosted on Codesphere
+
+Together with N8N, this template deploys an Ollama `qwen` model hosted on Codesphere. 
+To use it in N8N, add an `Ollama` node (from the AI category) and configure it: 
+
+```
+# Replace WORKSPACE_ID with your current Workspace ID!
+Base URL: http://ws-server-WORKSPACE_ID-ollama-service.workspaces.svc.cluster.local:3000
+API Key: <none>
+```

ci.yml

@@ -1,13 +1,41 @@
+schemaVersion: v0.2
 prepare:
   steps:
-    - name: install dependency
-      command: npm install cache-manager@5.2.4 --save
-    - name: install n8n
-      command: npm install n8n
+    - name: install Node.js 24
+      command: nix-env -iA nixpkgs.nodejs_24
+    - name: install Ollama
+      command: nix-env -iA nixpkgs.ollama nixpkgs.process-compose
+    - name: install dependencies
+      command: npm install
 test:
   steps: []
 run:
-  steps:
-    - name: deploy
-      command: sudo n install 20.0.0 && N8N_USER_FOLDER=/home/user/app/ N8N_PORT=3000
-        N8N_PROTOCOL=https ./node_modules/n8n/bin/n8n start --tunnel
+  n8n-frontend:
+    steps:
+      - name: n8n start
+        command: N8N_USER_FOLDER=/home/user/app/ N8N_PROTOCOL=http N8N_PORT=3000
+          N8N_HOST=ws-server-$WORKSPACE_ID-n8n-frontend.workspaces.svc.cluster.local
+          ./node_modules/n8n/bin/n8n start
+    plan: 8
+    replicas: 1
+    network:
+      ports:
+        - port: 3000
+          isPublic: false
+      paths:
+        - port: 3000
+          path: /
+          stripPath: false
+  ollama-service:
+    steps:
+      - command: process-compose -t=False -f process-compose.models.yaml
+    plan: 9
+    replicas: 1
+    network:
+      ports:
+        - port: 3000
+          isPublic: false
+      paths:
+        - port: 3000
+          path: /model
+          stripPath: true