Producer Checklist

Once the data steward approves your request via the Web Request Workflow, please complete the following items:

• Submit firewall requests
  • If this it the first DataHub service built for your target database (see existing services):
    • You need to submit any necessary Firewall Requests for connecting to your target database from DataHub:
      • Dev/Test: request access for source address 10.28.192.160/27
      • Prod: request access for source address 10.28.196.160/27
  • If DataHub services already exist for your target database:
    • You do not need to submit firewall requests for connecting to your target database from DataHub.
• Create Apigee Proxy
  • Download DataHub Starter Proxy for Apigee.
  • Download DataHub Documentation for Developers which includes step-by-step instructions on the Apigee setup on page 12-16.

Next, provide the following to the DataHub admins:

• SQL query
  • Remove newline characters and escape any double quotes with a backslash
• Configure as large query? (Yes or No)

What is a "Large" query in DataHub?

The AWS/Apigee services we use limit the response body size to 6MB and impose a hard time limit on returning results.

• If the query takes longer than 18 seconds it will not work with DataHub.
• If your query results exceed the 6MB limit on the payload but returns faster than 18 seconds we can configure it as a “large” query in DataHub: to get around the 6MB AWS payload size limit, DataHub returns a url the consuming application can use to download the data instead of returning the data directly in the response.
• This is a one-time decision from the outset; if you anticipate the results exceeding 6MB in the future, you should choose the "large" query option.

• Database connection information
  1. Connection string (including host, port, and service name)
  2. Database username
     • The system should be assigned a read-only DataHub user, with permissions for the tables that are needed
       • If this is the first DataHub service built for your target database: request a new read-only user be created for datahub, with the permissions needed for the tables in your query.
       • If DataHub services already exist for your target database: request that permissions for the tables needed in your query be added to the existing read-only datahub user.
  3. Database password
  4. Database type (e.g. oracle)
• Name for finished web service and one-line description.