TL;DR InfluxDB Tech Tips — Time Series Forecasting with Telegraf

This article discusses how software developers can use Telegraf Execd Processor Plugin for time series forecasting and anomaly detection. The author explains that Telegraf is a lightweight collection agent that can be configured with a single TOML configuration file to gather metrics from over 180 inputs and write data to various outputs and platforms. It also acts as a processor, aggregator, parser, and serializer. The Execd Processor Plugin makes Telegraf extensible in any language, allowing developers to process their data however they need. The author demonstrates how this plugin can be used to continuously generate forecasts and perform anomaly detection using Python scripts. The article also provides an example dataset from a beer brewing setup monitored by InfluxDB, highlighting the importance of maintaining constant temperature during fermentation. It discusses considerations for choosing an algorithm suitable for this purpose, such as requiring minimal data for training and generating accurate predictions for short forecast periods. The author explains that using the Telegraf Execd Processor Plugin offers several advantages, including being extremely responsive with new forecasts generated every time a point is written to InfluxDB. Additionally, it can run an external process once, saving startup costs compared to running scripts like cron jobs. Finally, the author encourages developers to take advantage of this plugin and contribute their own external plugins by opening a pull request on GitHub. They also invite feedback and assistance with any problems encountered while using or contributing to Telegraf.