Effective Data Localization Using Consistent Hashing in Cloud Time-Series Databases

Abstract

The commercial time-series database is suitable for processing the time-series data. However, a single commercial time-series database can only accommodate the time-series data acquired by limited amount of sensors. In this paper, in order to cope with the challenge of massive time-series data processing, we first propose a cloud time-series database framework based on commercial time-series databases, and then propose an effective consistent hashing based algorithm for solving the key problem, i.e., the data localization problem, in cloud time-series databases. A performance study shows the superiority of the framework and the algorithm for processing massive time-series data acquired by large amount of sensors.