A time-series database such as InfluxDB, TimescaleDB or Prometheus organizes data by series and time, which lets it append very large numbers of readings per second, compress similar consecutive values, and answer “average per hour over the last week” quickly. Built-in data retention and continuous downsampling keep storage from growing forever.
The choice matters at scale. A relational database works fine for a few hundred sensors, but at a few hundred thousand the write throughput and the size of indexes become the bottleneck, and the cost of keeping raw history becomes the dominant cloud bill. TSDBs trade general-purpose flexibility for this efficiency; joins with business data often happen elsewhere. See The IoT data pipeline.
Part of the IoT Glossary. Spotted an error or a missing term? Tell us.