Time Series Databases

Time Series Databases

Time Series Databases Explained



Time series data is a sequence of data points collected at regular intervals over a period of time. In short, it is any data that has a timestamp, including Internet of Things (IoT) device data, stocks, commodity prices, tide measurements, solar flare tracking, and health information. To provide value, this data requires aggregation and analysis. While the need to store time series data is not new, the dramatic growth of IoT- and sensor-based data has created scale, performance, fault tolerance, and high availability requirements that many databases cannot support. A time series database is optimized to meet the challenges of handling massive amounts of data from thousands or more devices.


The ability to scale out, up, and down as data grows is a basic requirement for time series use cases. NoSQL databases have emerged as a response to the inability of traditional relational databases to provide the scalability that is needed for many of today’s data needs. However, time series data presents additional, unique challenges that not all NoSQL databases are optimized to handle. When choosing a database for time series and IoT use cases, important considerations include:

  • Data location: If related data is not co-located on the same physical storage, queries can be slow and even result in time-outs. A database that co-locates a related time range of data on the same physical part of the database cluster will enable quick access for faster, more efficient analysis.
  • Fast, easy range queries: Analyzing time series data requires the ability to write range queries. With many databases, the sheer volume of time series data may cause the index to outpace the available memory, degrade read and write performance, and produce errors. A NoSQL database that co-locates related time series data ensures that requests complete without error. Additionally, consideration should be given to how easy or difficult the query language makes it for users to write queries.
  • High write performance: Many databases are not able to serve requests predictably and quickly during peak loads. Unless the database prioritizes cluster availability over strong consistency, the database may become less responsive or even completely unresponsive. A masterless, distributed NoSQL database is a good choice to ensure high availability and high performance for both read and write operations during peak loads because it is designed to stay available even under the most demanding conditions.
  • Data compaction: As time series data ages, the granularity of the data often becomes less relevant to the purpose of useful analysis. To be able to store and retrieve the data efficiently, organizations must have the ability to roll up and compress the data according to the needs of the business. A NoSQL database that facilitates the setting of appropriate data compaction levels will ensure ongoing fast read and write performance into the future.


A database that is optimized to handle time series data provides the following advantages:

Massive scalability and performance: An effective time series database enables an application to scale easily to support millions of IoT devices or time series data points in a continuous flow and perform real-time analysis.

Reduced downtime: In scenarios where downtime is unacceptable, the architecture of a database that is built for time series data ensures that data is always available even in the event of network partitions or hardware failures.

Lower costs: High resiliency translates into fewer resources needed to manage outages. Fast and easy scaling using commodity hardware reduces the operational and hardware costs of scaling up or down.

Improved business decisions: By enabling an organization to analyze data in real time, a time series database helps an organization make faster and more accurate adjustments for energy consumption, device maintenance, infrastructure changes, or other important decisions that impact the business.

tier-1-line class=


For organizations with enterprise-grade time series database requirements, Riak TS is optimized to deliver fast read and write performance along with massive scalability and resiliency.

Read about Riak TS


Time series is one of the biggest drivers of data-oriented applications today. Riak TS is engineered with purpose-built capabilities to address the critical requirements of handling time series data.

Read the blog post