Abstract:
A sensor data storage system is an important part of data analysis
systems. The duty of sensor data storage is to accept time series
data from remote sources, store them and provide access to
retrospective data on demand. As the number of sensors grows, the
storage system scaling becomes a concern. In this article we
experimentally evaluate Riak KV—a scalable distributed key-value
data store as a backend for a sensor data storage system (in Russian).
Key words and phrases:sensor data, write performance, distributed storage, time series, Riak, Erlang.