Another cruel lesson in the realities of multi-national cloud computing platforms. Did you hear that Microsoft’s Azure storage cloud fell down last week because of a school-boy error? Someone in Redmond Washington forgot to renew an SSL certificate and that took down the whole show. Every sub-region using the platform was reporting outages and degradation of service because of this issue. Mistakes happen. We are only human. But, shouldn’t you be able to trust the big guys to have the right infrastructure in place to prevent amateur mistakes like letting an SSL certifcate expire?
Forbes online contributor Tim Worstall summarizes the worry nicely,
…of course, things, mistakes, like this can happen to anyone. But the idea of handing over your entire computing system to a multinational running a cloud computing system is that they have in place the structures to stop such mistakes. … The entire point of the whole cloud computing enterprise is that you, the person wanting computer services, can get better services at lower cost by using the cloud. The point being that the cloud supplier can hire all the great engineers, gain economies of scale in hardware and bandwidth and so on. And that so on definitely includes dealing with all of the fiddly details of certificates. The point is to be able to take all of that stuff off the desk of some administrator in the purchasing company and have it done for all more efficiently by the supplier.
When I hear our industry lauding the cloud as the Panacea for all that ails your IT costs and complexity, I cringe … And this Azure failure is why. I’m not suggesting that ITW doesn’t make mistakes – of course we do. But when you have an issue on our watch, you have several people you can call day or night to tackle the problem and get you answers NOW. When the big cloud goes down (which it does) – who is your rep? Who can you call? Let me know if you find the phone number for Google or Microsoft support.
Caveat Emptor.
