Abstract:In IoT (internet of things), most data from the connected devices change with time and have sampling intervals, which are called time-series data. It is challenging to design a time series storage model that can write massive time-series data in a short time and can query and analyze the persistent time-series data for a long time. This paper constructs the RHTSDB (Redis-HBase Time Series Database) storage model based on Redis and HBase. RHTSDB uses the memory database Redis (Remote Dictionary Server) to cache massive time-series data, providing efficient data storage and query functions. HBase is used in RHTSDB for long-term storage of time-series data to realize their persistence. The paper designs a cold and hot separation mechanism for time-series data, where the infrequently accessed cold data are stored in HBase, and the frequently accessed and latest data are stored in Redis. Experiments verify that RHTSDB has apparent advantages over Apache IoTDB and HBase in data intake and query efficiency.

An Efficient NoSQL-Based Storage Schema for Large-Scale Time Series Data