Yandex Cloud Examples

This scenario demonstrates how to monitor the status of geographically distributed devices (e.g., vending machines) using Yandex Cloud serverless services. The architecture uses Yandex IoT Core for device connectivity, a PostgreSQL cluster for data storage, and Yandex DataLens for visualization.

Key components include:

• Yandex IoT Core: Receives MQTT messages from devices.
• Yandex Cloud Functions: Processes incoming data and emulates devices.
• Yandex Managed Service for PostgreSQL: Stores telemetry data.
• Yandex DataLens: Provides visual dashboards for monitoring.

This directory contains Terraform examples for deploying IoT scenarios within Yandex Cloud. The examples demonstrate how to set up infrastructure for device emulation and data ingestion.

Available scenarios:

1. Device Emulation: Uses Yandex Cloud Functions to emulate IoT devices (emulator_publish).
2. Data Ingestion to PostgreSQL: Uses Yandex Cloud Functions to subscribe to device data and write it into a PostgreSQL database (subscribe_and_postgresql).

This scenario demonstrates how to create a dashboard for monitoring sensor data and configuring alerts using Yandex Cloud serverless services (without virtual machines). It can be adapted for monitoring network equipment, industrial machinery, air quality, or climate control systems.

Data Model

Sensors transmit data via MQTT once per minute in the following JSON format:

{
    "DeviceId":"e7a68b2d-464e-4222-88bd-c9e8d10a70cd",
    "TimeStamp":"2020-05-21T10:16:43Z",
    "Values":[
        {"Type":"Float","Name":"Humidity","Value":"12.456"},
        {"Type":"Float","Name":"Temperature","Value":"-23.456"},
        {"Type":"Bool","Name":"Water sensor","Value":"false"},
        {"Type":"Bool","Name":"Smoke sensor","Value":"false"},
        {"Type":"Bool","Name":"Room door sensor","Value":"true"},
        {"Type":"Bool","Name":"Rack door sensor","Value":"false"}
    ]
}

This example demonstrates how to interact with a Yandex Cloud MQTT server using the Paho library. The example implements an 'echo' pattern: messages sent to the topic $devices/<device_id>/events are returned to the client via a subscription from the registry, which are then printed to the console.

Prerequisites

1. Create a Registry.
2. Create a Device within that registry.

Authentication Methods

The example supports two authorization methods:

• Certificates: Uses X.509 certificates for both the device and the registry.
• Login/Password: Uses the unique IDs of the registry and device as logins, paired with generated passwords.

The IoT scenarios in this repository use air sensor data as a source. The sensor outputs data in a JSON format containing a DeviceId, a TimeStamp, and an array of Values (e.g., Temperature, Humidity, Pressure, and CO2 levels).

{
 "DeviceId":"0e3ce1d0-1504-4325-972f-55c961319814",
 "TimeStamp":"2020-05-21T22:53:16Z",
 "Values":[
     {"Type":"Float","Name":"Humidity","Value":"25.281837"},
     {"Type":"Float","Name":"CarbonDioxide","Value":"67.96608"},
     {"Type":"Float","Name":"Pressure","Value":"110.7021"},
     {"Type":"Float","Name":"Temperature","Value":"127.708824"}
     ]
}

The project uses three isolated environments to manage the application lifecycle. Isolation is achieved at the IAM level, where each environment has its own directory in Yandex Cloud and its own set of static resources (databases, service accounts, etc.).

• prod: The production environment visible to end-users.
• testing: A testing environment used to validate the application before a release to prod.
• feature: A development environment used for creating temporary stands to debug and verify Pull Requests (PRs). Multiple stands can run simultaneously in this environment using a shared database.

Image Management: All Docker images are published to a shared infra directory. Service accounts for prod, testing, and feature have read-only access to this directory, while a dedicated builder service account is responsible for publishing images.