It’s kind of ironic to go through AWS well architected framework and add all the complexity associated with multi-region setups when Amazon themselves couldn’t get it right.

To be clear, I am not advocating for ignoring high availability setups. Just highlighting the complexity cost of it.