The second is that actually the big security vendors often care dramatically more than the customer about their data’s safety. Not only do they care more (because they have more on the line), but they have more resources than any normal business, even those in the Fortune 500, to put towards this problem.

The idea that big cloud providers do not provide 100% uptime “in a single datacenter” is not something that they even propose to offer. That’s a misunderstanding of the goals of those vendors and of cloud and only affects customers that either intentionally choose to be risky and forgo high uptimes or lacked IT understanding of designing their systems and were in no position to be designing those kinds of services. As these provides are not attempting to provide the kind of uptime that they are being asked to is a fault of customers, not the providers.

Thirdly, just because Amazon can’t produce 100% uptime doesn’t suggest that their security and uptime is not far better than the security and uptime of the customer. Small businesses cannot compete, at all, with the uptime of these providers. No service, none, provides 100% uptime. So ruling out the “better” service because it is not perfect is not logical.

Lastly, none of these points relate to colocation. The providers and types of providers and services being used as examples are very different things and not colocation providers.

Second, no one will ever care as much about a customer’s data or services as the customer itself. We have seen outages with Google, Amazon, Microsoft, and others. If they, with their nearly limitless engineering talent and financial resources, are unable to provide reliable service, it seems pretty risky to outsource IT to a third party whose system is beyond the customer’s control.