What is time series data and how is it used? - InfluxDB Tutorial

- [Instructor] Let's start this course by covering what is time series data and how it can affect data collection. This video will go over what time series data is and how it differs from other types of data that is out there. The definition of time series data is a sequence of data points that often consists of successive measurements made from the same source over a time interval. On this graph, you can see that each one of these points are successive measurements. These points are from the same data source over a period of time that is laid out on the X axis. Let's go over a couple real world examples that will help you better understand what is and isn't time series data. The first one is what I've always found fun is name popularity. I'm looking at the data for my own name, Samantha. In this chart, our Y axis is the fraction of that name for babies born that year with the X axis being that particular year. Here you can see Samantha peaked in the 1990s and has been on a downward trend over time. Financial data is another great example of time series data. This data is Bitcoin measured over a year. Each data point can be measured at a minute or even per second interval. Getting granular visibility and real time understanding of the exact time of a data point is very important in understanding the trends and volatility of cryptocurrency. Monitoring IoT data and building applications is one of the most common needs for a time series database. IoT data that users monitor will be constantly generated from hundreds to thousands of devices. This example shows energy creation from a handful of sources. Time series data produces both metrics and events. We define metrics as regular time series data where measurements are gathered consistently at regular time intervals. So this data can be collected every five seconds, every 10 minutes, or perhaps every 24 hours. These metrics could be an aerospace company collecting rocket ship data every 10 nanoseconds, or someone like me collecting GitHub statistics every day exactly at 8:00 AM. As opposed to metrics, events happen when they happen. The data is gathered at irregular time intervals. So for example, the data could be gathered at 10:02, then at 10:08, then not again until 10:30. For example, this could be stock data that only emits a data point every time the stock value changes. Events can be aggregated over a longer period of time which will then turn them into metrics. If we take our stock price example, we may have small fluctuations throughout the day, but the close price at the end of the day when the stock market closes is an aggregated time of the last reported value. And then the data then turns into a metric reported at a daily interval. So how do developers or application builders use time series data in their everyday life? Some common use cases are internal monitoring systems of infrastructure or applications to check in on its performance and any outages. Many time series users also use SaaS monitoring solutions to sell to customers. This could be manufacturing process monitoring or DevOps monitoring solutions. These are often cloud based and can be dropped in to be used by an array of users. And lastly, also time series data shows up in a lot of IoT products. These products, such as smart thermostats or electric charging vehicles will also include a custom application to help a user understand their time series data.