Time-series databases show much promise for developers looking to collect, visualize, process, and alert time series data. Yet several questions remain such as: why do you need a specialized database for time-series data, what should you consider when looking for a time-series database, and how to get started with a minimal amount of time and effort?
DBTA held a webinar with Al Sargent, senior director of product, InfluxData, who discussed how to evaluate the various time series databases on the market and get started fast.
According to Sargent, time series data is a sequence of data points, successive measurements made from same source over a time interval , and plot the points on a graph and your X axis is time.
For example, time series data includes weather conditions, stock prices, server monitoring, healthcare data, and more. Now, however, time series data includes server logs and traces, Sargent said.
Time Series databases are optimized for collecting, storing, retrieving, and processing of time series data. Primary use cases consist of:
- Industrial settings: factories, oil and gas, agriculture, smart roads and infrastructure
- Consumer: wearables, consumer devices, and trackers
- Custom monitoring solutions to track servers, VMs, applications, users, or events
- Apps that instrument business, social or development metrics in real-time
Sargent explained that time series databases require a system that can ingest metrics, events, logs, traces. A solution must also be able to accumulate, act, and analyze a broad range of data sources. It must also be able to gain broad enrichment from a range of relational databases.
Additional requirements include:
- Rapid ingest
- Built-in data exploration
- Built-in data processing
- Built-in data collaboration
- Choice of languages
- Deployment flexibility
- Built-in dashboarding
- Broad alerting
- Pre-built templates
- Outbound integrations
An archived on-demand replay of this webinar is available here.