Help: Ollama heartbeat configuration on Hostinger VPS (Docker)
Hi,
Setup:
• OpenClaw 2026.2.12 on Hostinger VPS (Docker)
• Ollama running in a separate Docker Compose stack (ollama-rsss)
• OpenClaw in its own stack (openclaw-bxnj)
• Connected both networks: docker network connect ollama-rsss_default openclaw-bxnj-openclaw-1
• Verified connectivity: curl http://ollama-rsss-ollama-1:11434/api/tags works from OpenClaw container
• OLLAMA_URL=http://ollama-rsss-ollama-1:11434 set in .env
• Model installed: llama3.1:latest (8B Q4_K_M)
Goal: Use Ollama/Llama 3.1 as heartbeat model (following the Token Optimization Guide by @mattganzak)
Config in openclaw.json:
"models": {
"mode": "merge",
"providers": {
"ollama": {
"baseUrl": "http://ollama-rsss-ollama-1:11434",
"models": [{"id": "llama3.1", "name": "ollama/llama3.1"}]
}
}
},
"agents": {
"defaults": {
"heartbeat": {"every": "5m", "model": "ollama/llama3.1"}
}
}
Auth profile added to auth-profiles.json:
"ollama:default": {"provider": "ollama", "type": "api_key", "key": "ollama-local"}
Errors I've seen:
1 No API key found for provider "ollama" (before adding auth profile)
2 Unknown model: ollama/llama3.1 (without providers config)
3 models.providers.ollama.models: Invalid input: expected array (when missing models array)
Question: What's the correct way to configure Ollama as a provider for heartbeat on a Docker deployment where Ollama runs in a separate stack? Is there a specific auth setup needed for local providers?
Thank you for your help
Setup:
• OpenClaw 2026.2.12 on Hostinger VPS (Docker)
• Ollama running in a separate Docker Compose stack (ollama-rsss)
• OpenClaw in its own stack (openclaw-bxnj)
• Connected both networks: docker network connect ollama-rsss_default openclaw-bxnj-openclaw-1
• Verified connectivity: curl http://ollama-rsss-ollama-1:11434/api/tags works from OpenClaw container
• OLLAMA_URL=http://ollama-rsss-ollama-1:11434 set in .env
• Model installed: llama3.1:latest (8B Q4_K_M)
Goal: Use Ollama/Llama 3.1 as heartbeat model (following the Token Optimization Guide by @mattganzak)
Config in openclaw.json:
"models": {
"mode": "merge",
"providers": {
"ollama": {
"baseUrl": "http://ollama-rsss-ollama-1:11434",
"models": [{"id": "llama3.1", "name": "ollama/llama3.1"}]
}
}
},
"agents": {
"defaults": {
"heartbeat": {"every": "5m", "model": "ollama/llama3.1"}
}
}
Auth profile added to auth-profiles.json:
"ollama:default": {"provider": "ollama", "type": "api_key", "key": "ollama-local"}
Errors I've seen:
1 No API key found for provider "ollama" (before adding auth profile)
2 Unknown model: ollama/llama3.1 (without providers config)
3 models.providers.ollama.models: Invalid input: expected array (when missing models array)
Question: What's the correct way to configure Ollama as a provider for heartbeat on a Docker deployment where Ollama runs in a separate stack? Is there a specific auth setup needed for local providers?
Thank you for your help
The AI that actually does things. Emails, calendar, home automation - all from your favorite chat app. New shell, ready to help. The lobster way.
129,194Members
Resources
Similar Threads
Was this page helpful?
