Syncing Azure AD groups is too slow

[AndreyMZ]

Problem description

The performance of syncing with Microsoft Entra ID (ex. Azure AD) groups added in https://github.com/DefectDojo/django-DefectDojo/pull/6128 is unacceptably poor. It can take 20 seconds to log in a user with 45 groups.

Root cause

1. https://github.com/DefectDojo/django-DefectDojo/blob/81c123e8d92024b965ddd3c985640e2fe398007b/dojo/pipeline.py#L80-L86

   To get the names DefectDojo requests Microsoft Graph REST API v1.0 Get group endpoint (GET /v1.0/groups/{id}) for every user’s group ID.

2. https://github.com/DefectDojo/django-DefectDojo/blob/81c123e8d92024b965ddd3c985640e2fe398007b/dojo/pipeline.py#L114-L120

   https://github.com/DefectDojo/django-DefectDojo/blob/81c123e8d92024b965ddd3c985640e2fe398007b/dojo/pipeline.py#L125-L130

   DefectDojo performs O(N) queries to the database, where N is a number of the user’s groups.

Possible solution

1. To reduce the number of network requests from N to 1 use the List transitive member of endpoint (GET /v1.0/me/transitiveMemberOf).
2. To reduce the number of database queries from O(N) to O(1) use bulk operations like:
   • in-bulk
   • bulk_create
   • RelatedManager.add
   • RelatedManager.remove