# Aggregator Pipelines

Aggregator pipelines load data through the aggregator instead of loading data directly to the leaves. Using aggregator pipelines can improve efficiency for low parallelism pipelines, such as single file S3 loads or single partition Kafka topics, and it is required for pipelines that load data into reference tables and tables with auto increment columns.

Aggregator pipelines only support stored procedures for MongoDB and MySQL sources, i.e., `CREATE AGGREGATOR PIPELINE ... AS LOAD DATA {MONGODB | MYSQL}` statements.

To create an aggregator pipeline, use the [CREATE AGGREGATOR PIPELINE](https://docs.singlestore.com/db/v9.1/reference/sql-reference/pipelines-commands/create-pipeline.md) command.

***

Modified at: February 4, 2025

Source: [/db/v9.1/load-data/about-singlestore-pipelines/pipeline-concepts/aggregator-pipelines/](https://docs.singlestore.com/db/v9.1/load-data/about-singlestore-pipelines/pipeline-concepts/aggregator-pipelines/)

(An index of the documentation is available at /llms.txt)