We send backend API trace data to apollo studio for analytics and to help monitor performance. This would be enhanced if we can tell which clients are creating the most traffic and when. At present all requests just group under the "Unidentified clients" category.
This should just need us to add an apollographql-client-name header to our requests (apollo docs). We could just set the header to something like "UOP frontend", but I wonder if it might be worth (if not problematic from a data protection stance) including some extra data, like a users id, to help identify when issues like https://github.com/UserOfficeProject/issue-tracker/issues/1094 pop up again?
What is the problem and why is it a problem
We send backend API trace data to apollo studio for analytics and to help monitor performance. This would be enhanced if we can tell which clients are creating the most traffic and when. At present all requests just group under the "Unidentified clients" category.
This should just need us to add an
apollographql-client-nameheader to our requests (apollo docs). We could just set the header to something like "UOP frontend", but I wonder if it might be worth (if not problematic from a data protection stance) including some extra data, like a users id, to help identify when issues like https://github.com/UserOfficeProject/issue-tracker/issues/1094 pop up again?Part of https://github.com/isisbusapps/ISISBusApps/issues/1006