The configurations table allows easier ingestion of subscans, without having to search for data descriptions.  Each execution block gets a configuration entry in this table, and the data descriptions are attached to this configuration as well. 

Definition:

       Column       |  Type   | Collation | Nullable |                         Default

--------------------+---------+-----------+----------+----------------------------------------------------------

 configuration_id   | integer |           | not null | nextval('configurations_configuration_id_seq'::regclass)

 configuration      | integer |           | not null |

 execution_block_id | integer |           | not null |

Indexes:

    "configurations_pk" PRIMARY KEY, btree (configuration_id)

Foreign-key constraints:

    "execution_blocks_configurations_fk" FOREIGN KEY (execution_block_id) REFERENCES execution_blocks(execution_block_id) ON UPDATE CASCADE ON DELETE CASCADE

Referenced by:

    TABLE "data_descriptions" CONSTRAINT "configurations_data_descriptions_fk" FOREIGN KEY (configuration_id) REFERENCES configurations(configuration_id) ON UPDATE CASCADE ON DELETE CASCADE

    TABLE "subscans" CONSTRAINT "configurations_subscans_fk" FOREIGN KEY (configuration_id) REFERENCES configurations(configuration_id) ON UPDATE CASCADE ON DELETE CASCADE

Columns:

configuration_id: auto-generated id.

configuration: used to connect back to the XML for VLA observations, or used to connect back to john's database during the import process for VLBA data.

execution_block_id: the id of the execution block that this configuration belongs to.