BigQuery is a serverless, fully managed, and petabyte-scale data warehouse solution for structured data hosted on the Google Cloud infrastructure. BigQuery provides an easy-to-learn and easy-to-use SQL-like language to query data for analysis. In BigQuery, data is organized as Tables, Rows, and Columns. BigQuery uses columnar storage to achieve high compression ratio and is efficient in executing ad hoc queries; the execution plans are optimized on the fly by BigQuery automatically. The reason BigQuery is capable of executing ad hoc queries is that it does not support or use any index, and the storage engine component of BigQuery continuously optimizes the way data is stored and organized. There are no maintenance jobs required to improve BigQuery's performance or clean up data to get better performance.

BigQuery can be accessed via a browser, command-line utility, or API. In this chapter, we will load data into a custom table via a browser by directly uploading the file to BigQuery and also importing data from a file in Google Cloud storage.

The hierarchy in BigQuery is Project | Datasets | Tables. Under a project, datasets can be created. Datasets are containers for tables. It is a way in which tables are grouped in a project. Tables belonging to different datasets in the same project can be combined in queries.