From the replication conference call, the requirement emerged that the gateway might need to store the URL of the THREDDS catalog as an access point of the top-level dataset in the catalog. Additionally, the metadata service needs to return this URL to the replication client.