> An other issue with that is when the service is used internationally / globally This also often happens prematurely. If the team in the US needs live data en

> An other issue with that is when the service is used internationally / globally This also often happens prematurely. If the team in the US needs live data entered in Asia then there's not much you can do, they system has to be global. But an often overlooked option (today) is running multiple instances of the same software with each only needing nine fives in it's region. Even if you do need live data it might be better to have another process shuffling it between international instances. Also helps with latency.