Customer Success Story: Home Power Usage Monitoring
Magee has been running an InfluxDB-based IoT system within his home for the past year as a proof of concept for his industrial applications. Through this test, he has been able to monitor his home climate control, electricity and natural gas use and analyze his heavy power and natural gas use habits.
The Telegraf MQTT Input Plugin is used to direct sensor data into the InfluxDB IoT bucket. In addition, the system and vSphere plugins are used to pull metrics from the edge server into a system bucket for monitoring the application health. As a further proof of concept, a secondary mechanism pushes a subset of the data to an InfluxDB Cloud database through the Google cloud pub/sub mechanism. Climate alerts indicating abnormally high or low temperatures are configured through the InfluxDB alert system. Dashboards for data review and analysis have been built using an edge-based Grafana server.
Magee likes InfluxDB because it integrates well with other tools, it’s open source, and it’s scalable. The open source aspect of the product works well with his Linux-based server infrastructure and allows him to spin up instances for testing and demonstration easily and without worry of license agreements. The scalability of the product allows applications to move from a proof-of-concept through to production with minimal re-work to scale up. The integration of the InfluxDB tools (Telegraf) with common IoT protocols means that integrating an InfluxDB system usually does not require additional applications to bring data from one system to another.
As an integrator who is more interested in providing solutions than products, InfluxDB offers Magee’s clients flexibility for their specific use cases whether it is locally hosted, centralized, or a hosted cloud implementation. This also offers them the ability to scale up an application as required. The input plugins developed for Telegraf offer a broad range of industrial application potential through MQTT, Modbus and OPC-UA. Using a native time series database for collecting process data is simply the right tool for the job and manages sensor-based data right out of the box.
From Magee’s experience, the best way to develop a solid application based on InfluxDB is to take advantage of the open source aspect and build, and to perform a number of end-to-end trials as you need to work out the best mechanisms for pulling data; the best mechanisms for reviewing data; and how to build your data fields to best align the output requirements with your available data. He recommends trying and failing until you have a robust application. He also recommends exhausting trials with the core InfluxDB components before introducing extra components and applications to achieve your goals.