For new data architects to have access to domain project, they must be added to the Google Group eco-data-architects@solvay.com.

All Data Architects have the ability to add, and remove people from this group.  

To grant a data architect explicit access to one project, add the user to the group of Data Architects that corresponds to the Domain.  See this page: Data Ocean Domain Access (Group Management)

The 1st prerequisite for this is that that a new GCP project has been provisioned for this Data Product.  See, question "I am starting as an architect on a new Data Product, how do I ask for the creation of a GCP Project?"





Before any investigation to understand why users are getting errors, it is imperative for users to produce an exact and exhaustive copy of the error message (either from BigQuery/Tableau/Dataiku/Qlik etc).  Most of the time, the reason is indicated in the error message.

Most errors will undoubtedly be related to inadequate permissions, either on the user side, or on the "server" side, or that the query itself is malformed.  Carefully examining the error will, in most cases, explain what the problem is.

In case of permission errors, it's necessary to verify the that the user is present in a group authorized to have access to the Data Product Project.  Normally the project manager is responsible for this verification.

Additionally, it may be that the user is querying a table or view in a dataset isn't shared with the group that the user is a member.  In this case, the dataset must be shared with the corresponding user group.



As a Data Architect, you are not directly responsible for the creation of GCP projects.  The request for new projects in GCP should be done by the Project Manager.  

This is documents on the wiki here: Data Product Project Request Information

However, as the Data Architect, you should work with the Project Manager to go over the template to define the necessary service accounts that will be required, as well as the groups that will be used in the project.

All GCP projects start with clean slate in BigQuery.



  • No labels