Back in 1993, Eric Schmidt, then the Sun kid, now the Google dad, wrote in an email to the telecosmic George Gilder: “When the network becomes as fast as the processor, the computer hollows out and spreads across the network.”
The Economist closed its recent article on cloud computing by sketching out a picture of where this technological trend is leading:
In future the geography of the cloud is likely to get even more complex. “Virtualisation” technology already allows the software running on individual servers to be moved from one data centre to another, mainly for back-up reasons. One day soon, these “virtual machines” may migrate to wherever computing power is cheapest, or energy is greenest. Then computing will have become a true utility—and it will no longer be apt to talk of computing clouds, so much as of a computing atmosphere.
Bill Thompson has noted that, as governments and corporations become more aware of, and either more nervous or more excited about, the ability to shift data and data processing effortlessly across borders, the “computing atmosphere” may get very foggy very fast, with the cloud turning “into a miasma … heavy with menace.” Through the noxious mist, Thompson can even hear hounds baying.
James Urquhart describes how the idea of the itinerant computer – a feather of software code wafting from data center to data center – is rapidly becoming, at a technical level, a reality:
The concept of “moving” servers around the world was greatly enhanced by the live motion technologies offered by all of the major virtualization infrastructure players (e.g. VMotion). With these technologies (as you all probably know by now), moving a server from one piece of hardware to another is as simple as clicking a button. Today, most of that convenience is limited to within a single network, but with upcoming SLAuto federation architectures and standards that inter-LAN motion will be greatly simplified over the coming years.
Once you’re able to “move your complete processing state from place to place as processing requires, without losing a beat,” a kind of legal arbitrage becomes possible:
So, run your registration process in the USA, your banking steps in Switzerland, and your gambling algorithms in the Bahamas. Or, market your child-focused alternative reality game in the US, but collect personal information exclusively on servers in Madagascar. It may still be technically illegal from a US perspective, but who do they prosecute? … I know there are a million roadblocks here, but I also know both the corporate world and underworld have proven themselves determined and ingenious technologists when it comes to these kinds of problems.
Gregory Ness suggests that the world’s new spice trails may be computing trails:
Over the last thousand or so years we’ve seen spice trails generate massive wealth in the Middle East, shipping lanes open up sizable agricultural and mining projects in less-developed regions; and steam, factories and electricity generate yet another wave of disproportionate winners. The wealth of North America in the last two decades has increasingly come from information technology and energy as manufacturing has chased low cost labor to nations with lower standards of living.
When spice trade routes shifted to the ocean the overall Middle East economy went from optimism to despair, from science and enlightenment to xenophobia. Factories gradually replaced artisans around the world and agriculture went through a series of cycles depending on access to trade routes and distances from markets (in addition to weather and practices, etc). A coming shift to cloud computing could be as influential in wealth distribution as any previous shift in factors of production and access.
Ness concludes: “It may only be a matter of time before we hear a politician talk about the evils of ‘cloudsourcing.'” For the moment, though, they’re celebrating in Lenoir, under an almost cloudless sky.
The government can choose to seize servers in its physical jurisdiction… or they could just follow the money to a person they can threaten. Which is, after all, their core competency :-)
So its worse than just server location. Control is important too. If you have a server in Frankfurt, a back office in Mumbai, and an HQ in London, there are three jurisdictions. Having physical access to a server is just one (inconvenient) point of control.
Consider a German hedge fund owning a sensitive stake in a Chinese bank, hosting servers in Switzerland, employing a Taiwanese national with joint British citizenship who visits the UK for meetings! You can make a good case for tracking corporate laptops and varying access as they move around…
(VMotion etc, don’t make much difference. They all rely on there being a shared storage pool replicated across the sites. This is the difficult bit. The economics of replicating a 15GB database across a continent are not great. Managing and replicating state across narrow links is a huge challenge.)
Thomas is correct in that moving 100s of GB of data even a few miles takes longer than folks expect, but smart design moves far less data, using replication. 15GB is not interesting, when terabyte disk drives sell for $200 at your local big box store.
The cypherpunks wrote about all of these issues in the mid-90s. When you setup virtual clouds of servers, you can have them be on “data havens” and then the rules of physical placement, and government control, become more of a challenge.
It will be interesting to watch where servers are going to be placed. Could be countries with cheap power, or well established legal protections, or perhaps isolated places where a few million dollars can convince the local politicians to ignore data extradition requests from the FBI and their equivalents.
There’s an interesting book by Edwin Black called IBM and the Holocaust which documents IBM’s strategic alliance with Nazi Germany proposing that it was IBM’s technologic assistance that allowed Hitler able to achieve the staggering numbers of the Holocaust.
It’s important to look at this link in an era in which people take it for granted that they have no right to privacy and that it is inevitable that there personal information leaves their personal control and moves into the cloud where anyone can access it and they have no control over it. To all those who are happy put there DOB, name address, religion, sexual orientation, friend lists and other personal information on Myspace, Facebook, PeopleData or whatever, take heed to what many of the holocaust survivors interviewed by Black remarked about the Gestapo when they showed up at their doors: “They always had names…..”
And “they” already have ours linuxguru.
The insidious creep, the “hounds” have already smelt blood, becomes the consensus trance that will most probably end with some pretty hefty jolts!
Aren’t Blackwater building camps already?
Alan
In the UK, sensitive people’s data is purged from databases (like HMRC’s) after the IRA used the data for targeting. It only takes 1 mole. What if Combat18 get a USB drive with everyone’s Facebook data? Not pleasant…
@ Patrick
Definitely, and the need to set replication up in advance stops functions from migrating arbitrarily with a click. (I’d disagree that 15GBs isn’t interesting. 15GBs is more than most company’s user and transactions data.)
— Sent from an Estonian coffee shop with Black Box Recorder playing in the background —
Se suscitan inevitablemente preguntas filosófico – económicas:
1. ¿Será ésta otra atmósfera también envenenable (o limpiable) por el hombre?
2. ¿Será su eminencia tal que llegará el día que no podamos “respirar” sin ella?
3. ¿Será ésta “gratis” como es hoy el aire sucio o habrá que pagar por ella pero a cambio será “aire limpio”?
Ventaja y desventaja de las metáforas: con-mueven el pensamiento, pero a la vez lo mueven demasiado lejos…
Imagine the orbital data center, financed by private, launched from Narvik, belonging to some Tonga robot. Will it be illegal to take it down from the high orbit by only suspiction in holding encrypted some “Al Quaeda 2048” data?
Fascinating piece Nick.
Been doing research on consumerism in healthcare and the move to allow consumers to take more direct control of their personal health information. Google & Microsoft have both released platforms with this in mind, but no one is really sure (at least from the consumer facing view) where exactly this data is being stored.
May not have the ramifications of an Al Queda, but health records are viewed as very sensitive information by consumers and it remains to be seen just how the cloud will provide protection. Is it a cloud with a silver lining or dark and ominous?