Complete guide to getting started with OpenFinOps
OpenFinOps is an open-source FinOps platform designed for AI/ML cost observability and optimization. It provides comprehensive visibility into your cloud and AI infrastructure spending with real-time dashboards, intelligent recommendations, and multi-cloud support.
OpenFinOps can be installed via pip, from source, or using Docker. The package supports Python 3.8 and above.
pip install openfinops
# Install with all cloud providers and features
pip install openfinops[all]
# Or install specific integrations
pip install openfinops[aws]
pip install openfinops[azure]
pip install openfinops[gcp]
git clone https://github.com/rdmurugan/openfinops.git
cd openfinops
pip install -e .
Note: For Web UI functionality, ensure you have flask-socketio and eventlet installed. These are included in the base installation.
Get up and running with OpenFinOps in under 5 minutes. This guide will show you how to launch the Web UI and start monitoring your costs.
# Using the CLI
openfinops server --host 127.0.0.1 --port 8080
# Or using Python
python -m openfinops server
Open your browser and navigate to:
http://localhost:8080/http://localhost:8080/dashboard/cfohttp://localhost:8080/dashboard/coohttp://localhost:8080/dashboard/infrastructureImportant: OpenFinOps uses agents that automatically discover resources and calculate costs.
from agents.aws_telemetry_agent import AWSTelemetryAgent
# Initialize AWS agent
agent = AWSTelemetryAgent(
openfinops_endpoint="http://localhost:8080",
aws_region="us-west-2"
)
# Register and run
if agent.register_agent():
print("✓ Agent registered")
# Collects metrics every 5 minutes
agent.run_continuous(interval_seconds=300)
The Web UI provides beautiful, real-time dashboards with WebSocket-based live updates. All dashboards feature glassmorphism design, Chart.js visualizations, and responsive layouts.
from openfinops.webui import start_server
# Start with custom configuration
start_server(
host='0.0.0.0',
port=8080,
debug=False
)
The Observability Hub receives and processes telemetry data from agents. Note: You don't call these methods directly - agents call them automatically.
Agent-Based Design: Deploy telemetry agents in your cloud accounts. Agents automatically discover resources, collect metrics, calculate costs, and send to the ObservabilityHub server.
from openfinops import ObservabilityHub
from openfinops.observability.cost_observatory import CostObservatory
# Initialize (server-side)
hub = ObservabilityHub()
cost_obs = CostObservatory()
# Query cluster health (populated by agents)
health = hub.get_cluster_health_summary()
for cluster_id, metrics in health.items():
print(f"{cluster_id}: {metrics['health_status']}")
# Query costs (populated by agents)
summary = cost_obs.get_cost_summary(time_range_hours=24)
print(f"Total 24h cost: ${summary['total_cost']:.2f}")
OpenFinOps provides role-based dashboards tailored for different stakeholders in your organization.
from openfinops.dashboard import CFODashboard
# Initialize CFO dashboard
cfo_dash = CFODashboard()
# Generate financial report
report = cfo_dash.generate_financial_report()
print(report.total_spend)
print(report.roi_analysis)
print(report.budget_status)
from openfinops.dashboard import InfrastructureLeaderDashboard
# Initialize infrastructure dashboard
infra_dash = InfrastructureLeaderDashboard()
# Get resource utilization
utilization = infra_dash.get_resource_utilization()
print(f"CPU: {utilization.cpu_percent}%")
print(f"Memory: {utilization.memory_percent}%")
Deploy telemetry agents as separate processes to automatically collect metrics and costs.
# deploy_aws_agent.py
from agents.aws_telemetry_agent import AWSTelemetryAgent
# Initialize agent (uses boto3 credential chain)
agent = AWSTelemetryAgent(
openfinops_endpoint="http://localhost:8080",
aws_region="us-west-2"
)
# Register with server
if agent.register_agent():
print("✓ Agent registered")
# Run continuous collection (every 5 minutes)
# Agent automatically:
# - Discovers EC2, EKS, Lambda, RDS, S3
# - Queries CloudWatch for metrics
# - Calculates costs from instance types
# - Sends to OpenFinOps server
agent.run_continuous(interval_seconds=300)
# Azure agent
from agents.azure_telemetry_agent import AzureTelemetryAgent
agent = AzureTelemetryAgent(
openfinops_endpoint="http://localhost:8080",
subscription_id="your-subscription-id"
)
agent.register_agent()
agent.run_continuous(interval_seconds=300)
# GCP agent
from agents.gcp_telemetry_agent import GCPTelemetryAgent
agent = GCPTelemetryAgent(
openfinops_endpoint="http://localhost:8080",
project_id="your-project-id"
)
agent.register_agent()
agent.run_continuous(interval_seconds=300)
Security Note: Always use IAM roles with minimal permissions when deploying telemetry agents. Never hardcode credentials.
OpenFinOps provides specialized agents for data platforms like Databricks and Snowflake, with automatic cost calculation based on DBUs, credits, and storage usage.
Track Databricks DBU consumption, cluster costs, job execution, and SQL warehouse usage.
# Install Databricks SDK
pip install databricks-sdk requests
# Set credentials
export DATABRICKS_HOST=https://your-workspace.cloud.databricks.com
export DATABRICKS_TOKEN=dapi***
# Run agent
python agents/databricks_telemetry_agent.py \
--openfinops-endpoint http://localhost:8080 \
--databricks-host $DATABRICKS_HOST \
--databricks-token $DATABRICKS_TOKEN \
--interval 300
Databricks Pricing:
Monitor Snowflake credit consumption, warehouse usage, storage, and query patterns.
# Install Snowflake connector
pip install snowflake-connector-python requests
# Set credentials
export SNOWFLAKE_USER=your_user
export SNOWFLAKE_PASSWORD=your_password
# Run agent
python agents/snowflake_telemetry_agent.py \
--openfinops-endpoint http://localhost:8080 \
--snowflake-account xy12345.us-east-1 \
--snowflake-warehouse COMPUTE_WH \
--edition enterprise \
--interval 300
Snowflake Editions:
Track costs from popular SaaS services including MongoDB Atlas, Redis Cloud, GitHub Actions, DataDog, and more.
The SaaS services agent can monitor multiple services from a single configuration.
# Create sample configuration
python agents/saas_services_telemetry_agent.py \
--create-config saas_config.json
# Run agent
python agents/saas_services_telemetry_agent.py \
--openfinops-endpoint http://localhost:8080 \
--config saas_config.json \
--interval 3600
{
"mongodb_atlas": {
"enabled": true,
"public_key": "your_public_key",
"private_key": "your_private_key",
"project_id": "your_project_id"
},
"redis_cloud": {
"enabled": true,
"api_key": "your_api_key",
"secret_key": "your_secret_key",
"account_id": "your_account_id"
},
"github_actions": {
"enabled": true,
"token": "ghp_your_token",
"org_name": "your_org"
},
"datadog": {
"enabled": true,
"api_key": "your_api_key",
"app_key": "your_app_key"
}
}
Current Support:
Coming Soon: Elasticsearch, Confluent Kafka, Vercel, Docker Hub
OpenFinOps can be configured via environment variables, configuration files, or programmatically.
export OPENFINOPS_HOST=0.0.0.0
export OPENFINOPS_PORT=8080
export OPENFINOPS_DEBUG=false
# Cloud provider credentials
export AWS_ACCESS_KEY_ID=your_key
export AWS_SECRET_ACCESS_KEY=your_secret
export AZURE_SUBSCRIPTION_ID=your_sub_id
export GCP_PROJECT_ID=your_project
# config.yaml
server:
host: 0.0.0.0
port: 8080
debug: false
observability:
update_interval: 5
retention_days: 90
alerts:
budget_threshold: 10000
anomaly_detection: true
notification_channels:
- email
- slack