China
Africa
Explore chapters and articles related to this topic
Intelligent Edge
Published in Haishi Bai, Zen of Cloud, 2019
Relational databases are designed for online transaction processing (OLTP), which expects random updates across scattered rows. IoT data presents a different operation pattern. Most IoT data is time series data, which means data points sorted by time. When these data points are inserted into a database, they are appended to the end of the table. This append-only operation pattern presents opportunities for optimization. For instance, using time as the primary index will make data insertion and query more efficient in this case. However, as a table holds data from multiple sensors, data points from different sensors will be mingled together. A way to solve this is to partition the database by devices. However, this might be impractical if you have many devices or require many cross-device queries.
One Platform Rules All
Published in Kuan-Ching Li, Hai Jiang, Albert Y. Zomaya, Big Data Management and Processing, 2017
Many researchers put their efforts into developing working database management systems, which are completely based on the new relational model. Two of the works are prominent: one is Ingres from University of Californian at Berkeley and the other is System R from IBM. In 1974, IBM engineers invented a query language named SEQUEL (structured English query language) for System R, which was predecessor of the SQL language. Besides that, researchers have developed storage methods, indexing techniques, query optimization techniques, transaction-processing methods, and database recovery techniques for RDBMS. These methods made RDBMS an efficient and reliable data engine for various online transaction-processing (OLTP) applications.
A Review of Spatial Big Data Platforms, Opportunities, and Challenges
Published in IETE Journal of Education, 2020
Since the 1970s, the Relational Database Management System (RDBMS) is used in Online Transaction Processing (OLTP) systems such as banking, airline reservation system, payroll, and more for storage, processing, and querying of data in text and numeric form [12]. These systems are developed based on client server architecture with databases at the backend. The RDBMS doesn’t provide native support for spatial attributes of type point, line and polygon, and operations like distance, intersect, and near performed on them. The workaround in RDBMS for storing vector data like polygon models the location data in alphanumeric form and stores them across multiple tables. Although these workarounds can help store spatial data in RDBMS, joining and querying these data to perform spatial operations is inefficient and expensive.