DAO End-to-End Sample Flow

Configure devices to automatically propose and vote based on collected metrics.

Overview

The “OASEES agent” (collector + blockchain agent) can:

• Propose actions based on metric conditions
• Vote on proposals based on metric thresholds
• Execute actions when proposals pass

Configuration File

Generate Template

Create a configuration template:

oasees-sdk telemetry gen-config

This generates config.json with the following structure:

{
    "metric_index": " ",
    "propose_on": {
        "events": [" "],
        "proposal_contents": [" "],
        "positive_vote_on": [" "]
    },
    "actions_map": {
        " ": {
            "action_endpoint": "",
            "args": {}
        }
    },
    "leader": ""
}

Configuration Fields

Field	Optional	Description
metric_index		Name of the metrics index (e.g., swarm_metrics)
propose_on.events		Conditions that trigger a proposal
propose_on.proposal_contents		Message and action value for the proposal
propose_on.positive_vote_on		Conditions for voting YES
actions_map		Maps DAO action values to actual API calls
leader	X	Pick a specific device to execute the API call when the proposal is executed. The default behavior is that all devices execute the API call

The three propose_on subfields are lists with one-to-one binding. They must have equal numbers of elements. The 2nd event binds to the 2nd proposal_content and 2nd positive_vote_on condition.

Example Configuration

{
    "metric_index": "swarm_metrics",
    "propose_on": {
        "events": [
            "thermal_value > 99",
            "pressure_value > 60"
        ],
        "proposal_contents": [
            {
                "msg": "High Temperature, cool-down needed",
                "action_value": 1
            },
            {
                "msg": "High Pressure, emergency",
                "action_value": 2
            }
        ],
        "positive_vote_on": [
            "flow_value > 7",
            "thermal_value < 50"
        ]
    },
    "actions_map": {
        "1": {
            "action_endpoint": "http://localhost:5000/send_message",
            "args": {
                "target_sensors": [
                    "thermal-sensor",
                    "pressure-sensor",
                    "flow-sensor"
                ],
                "action": "cool_down",
                "duration": 10
            }
        },
        "2": {
            "action_endpoint": "http://localhost:5000/send_message",
            "args": {
                "target_sensors": [
                    "thermal-sensor",
                    "pressure-sensor",
                    "flow-sensor"
                ],
                "action": "emergency",
                "duration": 15
            }
        }
    },
    "leader": "device_name"
}

Configuration Interpretation

Scenario 1: High Temperature

Trigger: Device detects thermal_value > 99

1. Proposes: “High Temperature, cool-down needed” with action_value: 1
2. Voting: Devices with flow_value > 7 vote YES, others vote NO
3. If Passed: Execute action mapped to "1":
   • POST to http://localhost:5000/send_message
   • With args: action: "cool_down", duration: 10

Scenario 2: High Pressure

Trigger: Device detects pressure_value > 60

1. Proposes: “High Pressure, emergency” with action_value: 2
2. Voting: Devices with thermal_value < 50 vote YES, others vote NO
3. If Passed: Execute action mapped to "2":
   • POST to http://localhost:5000/send_message
   • With args: action: "emergency", duration: 15

Deploying Configuration

Apply Configuration to Agents

oasees-sdk telemetry configure-agents config.json

This command:

1. Reads your configuration
2. Deploys it to all blockchain agents
3. Agents begin monitoring metrics and responding accordingly

Action Endpoints

Action endpoints must be POST endpoints that accept JSON payloads.

Endpoint Requirements

• Use-case specific: Implement logic for your application
• Accessible: Must be reachable from Kubernetes pods
• POST method: Must accept POST requests with JSON body
• Localhost allowed: Endpoints can use localhost since agents run in Kubernetes

Example Endpoint Structure

@app.route('/send_message', methods=['POST'])
def send_message():
    data = request.get_json()
    target_sensors = data.get('target_sensors', [])
    action = data.get('action', 'stabilize')
    duration = data.get('duration', 30)

    # Implement your action logic here
    # ...

    return jsonify({'status': 'success'})

Testing the Workflow

Force Metrics for Testing

For the sample application, use the provided utility script:

import requests

# Force metrics to trigger DAO flow
response = requests.post(
    'http://localhost:5000/force_metrics',
    json={
        'value': 42.5,
        'duration': 60
    }
)

print(response.json())

Change localhost:5000 to your actual sensor endpoint.

Monitoring the Flow

1. Check Metrics: View metrics in the portal or via:

   oasees-sdk telemetry metrics-list -i swarm_metrics
   

2. Watch Proposals: Monitor DAO proposals in the portal

3. Verify Voting: Check that devices are voting based on their metrics

4. Confirm Actions: Verify actions are executed when proposals pass

Advanced Configuration

Complex Proposal / Voting Logic

Combine multiple metrics in conditions:

• Same device metrics: Use and/or operators (and takes precedence)
• Different device metrics: Separate with comma (,)

Example

• Setup: The sample application with 2 devices as participants.
  • Device1 is responsible for the flow sensor
  • Device2 is responsible for the thermal and pressure sensors.
• Conditions:
  • Device1 votes yes on: flow_value > 10
  • Device2 votes yes on: thermal_value > 80 && pressure_value < 20
• Config syntax:

  "positive_vote_on": [
    "flow_value > 10, thermal_value > 80 and pressure_value < 20",
  ]
  

If a device’s metric isn’t included in the voting conditions, the current default behavior is to vote AGAINST.

Troubleshooting

Proposals Not Being Created

• Verify collector is ingesting metrics
• Check metric values meet event conditions
• Ensure configuration was applied: oasees-sdk telemetry configure-agents
• Review agent logs

Actions Not Executing

• Verify endpoint is accessible
• Check endpoint accepts POST with JSON
• Ensure action_map value matches DAO action value
• Review application logs

Voting Not Working

• Verify devices have vote tokens
• Check voting conditions match actual metric values
• Ensure DAO quorum is achievable
• Review blockchain transactions

Related Video

• 6.configure-agents-DAO-flow.mp4

Next Steps

Back to DAO Operations

Continue to MLops & Federated Learning