These type of threads always end up being "cloud vs dedicated" when in my opinion it should be about the architecture. It shouldn't matter if you are running de

These type of threads always end up being "cloud vs dedicated" when in my opinion it should be about the architecture. It shouldn't matter if you are running dedicated vs cloud assuming a fairly reliable provider. It should be about avoiding single points of failure and redundancy. eg. If there is a hardware failure, can the site still remain operational? Of course resources cost money, but you could have 2 cheaper/smaller servers load balanced for around the price of 1 pricier/bigger server.