The serverless Presto engine is a Presto-based engine that is used for interactive analysis. Presto was designed to cope with time-consuming online analysis by using Hive. Presto uses an in-memory execution engine. Compared with other engines that write intermediate data to disks, Presto provides faster computing and execution. Presto is especially suitable for data analytics workloads, such as ad hoc queries, BI analysis, and lightweight extract, transform, and load (ETL).

Data Lake Analytics (DLA) is developed based on Presto. DLA supports most of Alibaba Cloud data sources, such as AnalyticDB for MySQL, AnalyticDB for PostgreSQL, and Tablestore. DLA uses the optimized Hive connector to significantly reduce the number of Object Storage Service (OSS) calls during OSS data analysis. This improves analysis performance and reduces costs. DLA has an enterprise-level access control system to protect your data. DLA also provides a high-availability coordinator to improve the overall service availability. DLA supports the MySQL protocol so that you can use MySQL-compatible tools for data analysis.

The following figure shows the architecture of the serverless Presto engine.
  • FrontNode provides user access and supports the MySQL protocol. You can use MySQL-compatible clients to access MySQL services provided by the serverless Presto engine.
    • In the DLA console, you can create databases and tables and query data on the Execute page.
    • You can also use other clients to access services provided by the serverless Presto engine. For more information, see Create an endpoint.
      Note To connect DLA to a database by using Data Management (DMS), you must enter the database account and password.
  • Presto clusters created in DLA are used for data analysis and computing. You can deploy multiple clusters in each region. By default, one cluster is charged based on the number of bytes scanned, and other clusters are charged based on the number of CUs consumed.
    • For more information about the differences between billing methods, see Billing methods.
    • For more information about the differences between the edition that is charged based on the number of bytes scanned and the edition that is charged based on the number of CUs consumed, see Differences between billing methods.
    • For more information about how to activate and use DLA Presto CU Edition, see Use the DLA Presto-compatible Presto CU edition.
    • For more information about SQL statements for creating databases and tables, querying data, and managing permissions, see SQL reference. The database and table creation options vary based on data sources. For more information about how to create databases and tables for each data source, see Access data sources.
    • Presto clusters created in DLA are compatible with open source Presto clusters. For the definition of functions supported by open source Presto clusters, see PrestoDB Functions and Operators.
  • Presto clusters created in DLA can access various data sources.