Skip to main content

Connections - Beta

Connections allow you to import data into Macrometa Global Data Network (GDN) from external data sources or other Macrometa document collections. You can also export data from Macrometa GDN into external data sources or other Macrometa document collections.

Once created, you can reuse a connection as many times as necessary.


This feature is currently in beta. Contact if you want to try it out.

Data Source, Connector, Workflow

These terms are used frequently when discussing connections.

A data source refers to an external database or entity that contains data, such as MongoDB, PostgresSQL, a different Macrometa collection, and so on.

A connector is a connection that can be set up between Macrometa and a data source. For example, the list of connectors includes MongoDB, Snowflake, Macrometa collections, and more.

An connection is a connection that you have created using the connector between Macrometa and a specific data source or target.

An ETL workflow is a connection between a specific collection and a specific table in a data source.

Sources and Targets

A source connection allows you to use an external data source or a Macrometa collection as the originating source for data. Any records entered in the source are sent to a Macrometa collection that you choose. A collection can only have one source, and it must be set at collection creation.

A target connection allows you to send data from a Macrometa collection to an external data source or a Macrometa collection. Any records entered in the collection are sent to the target that you choose. A Macrometa collection can have multiple targets.


You can transform data moving in or out of Macrometa, including creating roll-ups. For more information, refer to Transformations.

Common Schema

If a field has more than one data type, then the data type that applies to the majority of the first 50 records will be selected for import. The rest will be ignored.

Updating Source Tables

Once a log-based source connection is created, then Macrometa listens to inserts, updates, and deletes. It does not recognize table definition changes, such as changing a column name.

If you alter a source table, such as removing or renaming a column, after it has been mapped to a Macrometa collection, then records will not be updated. You must reload the collection in order to update the schema.

For more information about how Macrometa ingests data from sources, refer to Ingestion Modes.


Here are some limitations that you should consider as you work with collections and connections:

  • You cannot change a source connection assigned to a collection after the collection is created.
  • You cannot change a connection name.
  • An empty collection cannot be a target for an connection.
  • You cannot edit transformations.
  • Play tier users, refer to Tenant Quotas and Limits to view limits on number of connections. Keep in mind that each connection workflow also adds one collection, and each transformation adds one stream worker.
  • When you add a target, you only capture subsequent inserts, updates, and deletes made on the data after the connection is made. Truncating the collection does not count as deleting records in this case.
  • To support actions such as deleting records using workflows, Macrometa transfers additional metadata such as _sdc_deleted_at.
  • Metadata size is included in metrics calculations retrieved using the Retrieve Metrics API. This means that metric sizes such as exported_bytes might not be identical to the total size of the data in the source collection.