Unify and manage your data

Data Loader API upload requirements

Learn about requirements for uploading different file types and sizes in the Reltio Data Loader API.

The Data Loader has different requirements for uploading various file formats.

Important: We limit the number of records in a single file to a maximum of ten million records. This limitation applies to all supported formats. For files with large data amounts, we recommend that you split the load into batches with no greater than one million records each.

Supported file formats and sizes

We do not check file size for remote files. The limit is ten million rows per file, regardless of file size. The Data Loader doesn't have checks for file size. We verify row count and number of objects. Follow the requirements for the data file format that you want to upload.
File formatRequirements
CSV (comma-separated value)
  • Remote files have a limit of 10 million rows per file, regardless of file size.

  • Local file uploads have a maximum size limit of 50 megabytes.

For general CSV requirement specifications, see topic RFC 4180 - Common Format and MIME Type for Comma-Separated Values (CSV) Files.

Microsoft Excel
  • Remote files have a limit of 10 million rows per file, regardless of file size.

  • Local file uploads have a maximum size limit of 50 megabytes.

JSON

Remote and local files have a limit of 10 million rows per file, regardless of file size.

Note: The allocation remains the same regardless of the location where you’re uploading from (local file system or remote storage).

For information about other Data Loader requirements, see topic Get ready for Data Loader.

Trim special characters

Before data is loaded into Reltio, we ensure that the following special characters are trimmed:

  • Leading and trailing spaces: ' '.
  • Leading and trailing tabs: \t.
  • New line: \n.
  • Carriage return: r.

Trimming special characters does not apply to RELTIO_JSON files, since unlike a normal JSON file, the RELTIO_JSON file contains the information needed for the data upload without requiring additional table transformation.