If all it took were millions of dollars and slightly harder thinking to make a completely failure-proof system, a lot of problems would suddenly and cheaply be solved. At the end of the day there's still hardware, software, network infrastructure, peering, random chance, and all of the imperfections in the humans designing these systems.

> definitely not suppose to happen

Over what time range are you referring? Based on the 'millions' comment, I'm assuming you mean 'all time'. It's illogical to expect systems to perform flawlessly over a long period of time.

I'd also quantify the fact that it was only "unavailable" for a period. It's not like they lost my data.

If government teaches us anything, it's that there's no correlation between money spent and reliability of the system. :/

No, I just think that you'd be a fool to have ever believed any promise that the cloud would eliminate downtime. Even if from a technical perspective that were true, there's still the possibility of human error.