Overview

These pages describe popular datasets that you can load into YDB to familiarize yourself with the database's functionality and test various use cases.

Prerequisites

To load the datasets, you will need:

  1. Installed YDB CLI
  2. [Optional] Configured connection profile to YDB to avoid specifying connection parameters with every command

General Information on Data Loading

YDB supports importing data from CSV files using the command ydb import file csv. Example command:

ydb import file csv --header --null-value "" --path <table_path> <file>.csv

Where:

  • --header indicates that the first row of the file contains the column names, and the actual data starts from the second row;
  • --null-value "" specifies that an empty string in the CSV will be interpreted as a null value during data import into the table.

A table must already exist in YDB for data import. The primary way to create a table is by executing the CREATE TABLE YQL query. Instead of writing the query manually, you can try running the import command from a file, as shown in any example in this section, without creating the table first. In this case, the CLI will suggest a CREATE TABLE query, which you can use as a base, edit if necessary, and execute.

To import data into YDB, a table must be pre-created. Typically, a table is created using the YQL CREATE TABLE query. However, instead of crafting such a query manually, you can initiate the import command ydb import file csv as shown in the import examples in this section. If the table doesn't exist, the CLI will automatically suggest a CREATE TABLE query that you can use to create the table.

Selecting a Primary Key

YDB requires a primary key for the table. It significantly speeds up data loading and processing, and it also allows for deduplication: rows with identical values in the primary key columns replace each other.

If the imported dataset doesn't have suitable columns for a primary key, we add a new column with row numbers and use it as the primary key, as each row number is unique within the file.

Features and Limitations

When woimporting data to YDB, consider the following points:

  1. Column Names: Column names should not contain spaces or special characters.

  2. Data Types:

    • Date/time strings with timezone (e.g., "2019-11-01 00:00:00 UTC") will be imported as Text type.
    • The Bool type is not supported as a column type; use Text or Int64 instead.

Available Datasets

Previous
Quick start
Next
Overview