NagareDB: A Resource-Efficient Document-Oriented Time-Series Database

The recent great technological advance has led to a broad proliferation of Monitoring Infrastructures, which typically keep track specific assets along time, ranging from factory machinery, device location, or even people. Gathering this data become crucial for wide number applications, like exploration dashboards Machine Learning techniques, such as Anomaly Detection. Time-Series Databases, designed handle these data, grew in popularity, becoming the fastest-growing database type 2019. In consequence, keeping and mastering those rapidly evolving technologies became increasingly difficult. This paper introduces holistic design approach followed building NagareDB, built on top MongoDB—the most popular NoSQL Database, discouraged scenario. goal NagareDB is ease access three essential resources needed time-dependent systems: Hardware, since it able work commodity machines; Software, an open-source solution; Expert Personnel, its foundation considered DB, lowering learning curve. Concretely, outperform MongoDB recommended implementation up 4.7 times, when retrieving while also offering stream-ingestion 35% faster than InfluxDB, database. Moreover, by relaxing some requirements, reduce disk space usage 40%.