Please fill in the following information:

Issue description

Improvement Feedback

=====================

Description:  ch04s05
We are using noobaa for logging, without using any mcg relevant features.

The chapter needs to explain why noobaa has been choosen instead of rgw.

I don't see the reason, since rgw is an actual s3 endpoint and it can, if an se is requested, become fault tolerent, which noobaa can't.

imo. if we should use noobaa here, then we need to introduce a requirement to replicate the objects to external s3 endpoints, in addition to having the data locally, thereby leveraging the actual benefit of noobaa and countering some of the inherent drawbacks. 

Steps to reproduce:

Workaround:

Expected result: