In “The hidden risk of a meltdown in the cloud,” a Technology Review blogger reacts to a paper by Bryan Ford on “the unrecognised risks of cloud computing.” I don’t know, the risks seem familiar to me. Beyond security, they are:
Unpredictable interactions among loosely-coupled services
Inability to preserve or reproduce an application or data set
The Technology Review blogger, who is evidently known only by the nom de plume Kentucky FC, echoes Ford’s conclusion: We ought to study these risks “before our socioeconomic system becomes completely and irreversibly dependent on a computing model whose foundations may still be incompletely understood.”
OK, yes, we should study the risks. But that doesn’t mean we can’t engage with the cloud while doing so. It isn’t an all-or-none proposition.
Think about our relationship to the power grid. We are, in fact, irrevocably committed to it. And it is prone to occasional dramatic failures. I have a few friends who live off the grid, but most of us plug in, and then some hedge their bets to varying degrees. Do you own a generator? If so, how much of your demand does it power? And for long? An hour? A day? A week?
For enterprises, a hybrid strategy that blends cloud and on-premise resources is gaining traction. That’ll make sense for individuals too. Our personal clouds encompass resources both in the sky and scattered across our own devices. As we extend into the cloud we’ll learn how to use it to complement the strengths and offset the weakness of our local setups. There is, as always, a continuum of risk and benefit. We’ll make personal choices to occupy points along that continuum. And those points will drift over time.
Meanwhile, let’s consider one analogy drawn by Bryan Ford and echoed by Kentucky FC.
Ford: Non-transparent layering structures … may create unexpected and potentially catastrophic failure correlations, reminiscent of financial industry crashes.
KFC: The cloud could suffer the same kind of collapses that plague the financial system….
It’s true that the unpredictability of complex interaction is a similar concern in both realms. But when things have gone wrong, cloud providers have been refreshingly open about it. Consider the post-mortems for some notable Amazon Web Services (AWS) and Azure outages. Both set a high standard, explaining what went wrong, why, how it was fixed, and what steps are being taken to prevent a recurrence.
We can only dream of a financial industry that runs as transparently, and holds itself to such a standard.