IBM Cloud Docs
Presto (C++)

Presto (C++)

Presto (C++) is a version of Presto workers that are implemented in C++ instead of Java by using the Velox library.

Presto (C++) aims to enhance performance for data lakes without requiring a JVM on worker nodes. It supports several connectors, including Hive and Iceberg, and focuses on improved integration with data warehousing systems.

For more information about provisioning the engine, see Provisioning a Presto (C++) engine.

Presto (C++) features

  • Task management: Presto C++ includes HTTP endpoints that allow users to monitor and manage tasks. This feature enhances operational oversight and makes it easier to track ongoing processes.
  • Remote function execution: Enables executing functions on remote nodes, which enhance scalability and distributed processing capabilities, making data processing more efficient across a network of nodes.
  • Authentication: Uses JSON Web Tokens (JWT) for secure internal communication between nodes, ensuring that data remains secure and tamper-proof during transmission.
  • Data caching: Implements asynchronous data caching with prefetching capabilities. This optimizes data retrieval and processing speed by anticipating data needs and caching it in advance.
  • Performance Tuning: Offers various session properties for performance tuning, including settings for spill thresholds and compression. This allows users to fine-tune performance parameters according to their specific needs, ensuring optimal performance of data processing tasks.

For more information about Presto (C++), see Presto C++ features.

For more information about provisioning a Presto (C++) engine, see Provisioning a Presto (C++) engine.

For more information about customizing properties, see: