Excessive calls to AWS get_role_tags; region awareness

[achantavy]

Description:

What issue is being seen? Describe what should be happening instead of the bug, for example: Cartography should not crash, the expected value isn't returned, the data schema is wrong, etc.

The AWS resourcegroupstaggingapi sync right now runs once per region, which makes sense for resource discovery since most of these resources are region-specific.
However iam roles are global (since the arn of a role does not have a region in it), so our workaround to call get_role_tags will actually be called once per region, which makes things quite slow (a full run of us-east-1 actually timed out on my laptop's boto3 session).

What we should do

1. We should find a way to avoid wasting API calls to get_role_tags. Maybe this can involve storing or calculating information about which resources are global versus regional.
2. It does not make sense for AWSTag nodes to have a region as one of their properties so we should remove region from the node.

Please complete the following information::

• Cartography release version or commit hash [e.g. 0.12.0 or 95e8e11]

0.72.1