compare both lists, and only post users that either were in auth0 but marked as blocked (ie to be disabled) or users that were not present in the CIS list. This is to save time on signing, processing, posting of course. There are normally no changes that we actually care about at this point. When we care about changes, we need to pull in the CIS profiles (not just the user_ids) which can be expensive.
post the result using the same system as the hris_publisher, which is, fan out to multiple lambda, and the posting itself is also threaded within the lambda function
This publisher will:
blocked(ie to be disabled) or users that were not present in the CIS list. This is to save time on signing, processing, posting of course. There are normally no changes that we actually care about at this point. When we care about changes, we need to pull in the CIS profiles (not just the user_ids) which can be expensive.what this is missing: