Connecting external catalogs

The external catalog connection feature is designed to integrate data from external sources into Tengri. Tables from external sources can be replicated to the data store within Tengri, or you can access them directly via a connected catalog without replication.

To launch the external catalog connection wizard, click the Add Catalog button on the Catalog tab, select a connector (the following options are currently available: PostgreSQL, HiveMetaStore) and follow the further instructions.

Specify a catalog name for the source you are connecting. This name will be displayed in the left-hand side menu in the list of connected catalogs (alongside the system TNGRi Catalog).

Table replication

Once an external catalog has been connected, you can, if necessary, replicate tables from the connected catalog to the storage within Tengri. To do this, you must select the list of schemas and/or tables to be replicated, the target schema in the main catalog (where the replicated tables will be created) and a mask for the names of the replicated tables.

Tengri interface for working with external catalogs

By default, the mask for the names of the replicated tables is: {catalog}__{schema}__{table}.

Querying the replicated tables in SQL is no different from querying other tables in the main catalog. For example, if tables have been replicated to the raw schema and the default table name mask has been used, a query against the table public.table1 from the external catalog my_catalog would look like this:

SELECT * FROM raw.my_catalog__public__table1;

Accessing tables from external catalogs

To access tables directly from connected external catalogs, use the external catalog prefix. For example, a query to the same table would look like this:

SELECT * FROM my_catalog.public.table1;

For convenience, the full table names (from both external catalogs and the main one), including all prefixes, are listed in the header of the view page for the relevant table in the catalog.