Efficient Retrieval of Parquet Data | Native Object Store | Teradata Vantage - 17.10 - Using Filtering to Retrieve Parquet Data More Efficiently - Advanced SQL Engine - Teradata Database

Teradata Vantage™ - Native Object Store Getting Started Guide

Advanced SQL Engine
Teradata Database
Release Number
Release Date
July 2021
Content Type
Programming Reference
Publication ID
English (United States)

Reading all the objects in a bucket can be time-consuming and expensive. An alternative solution is to filter the object store using one or both of the following methods:

Path Filtering
  • A path is a series of values that constitute a multi-level key (siteno, year, month, day) which identifies one or more objects within an object store.
  • Path filtering lets you pre-select which objects in the object store you want to bring back to the database, by providing specified values for different levels in a path key (such as "WHERE $path.$siteno = 09380000")
Column Filtering
  • Parquet supports the ability to apply filtering at the column level while the data is in the object store.
  • This can be as effective as path filtering in reducing the volume of data returned.