Cloud Computing Still Too Exciting; Feds, Rackspace Try to Dull It Up a Little
'Credible,' 'reliable,' 'exciting,' 'unpredictable' don't all belong in descriptions of IT infrastructure
An important element in credibility is a certain level of tedium. Especially in technology, if something is really exciting, the odds are that it's too new or too ambitious to be really good at what it does, or at least to be really reliable.
That may be less true of gaming tech or flashy Web sites or cool new information services.
When it comes to anything that could remotely be considered to be infrastructure, being boring is at least as important as any other element that goes into it.
An infrastructure product that is boring performs its function well, everything that happens when it is used is predictable, as are the things that happen afterward and the beginning of the next cycle.
It's exciting in an auto plant to see a door-panel or other part stamped out of steel by giant hydraulic hammers, because they're huge and loud and heavy. After the 40th or 50th cycle through the stamper, though, if it's still exciting it's probably because pieces of doorpanel are flying all over the place or Stan the stamper guy got too close and things got all CSI, or some other horrible thing happened.
Technology that's credible to the people who sign the checks is technology that does what it's supposed to every time, without extraordinary measures from its minders from ITor extra-special setups so it won't screw up.
Two things happened today that will add quite a lot of tedium to world of cloud computing:
First, the U.S. General Services Administration gave permission for 11 companies to offer infrastructure-as-a-service to federal agencies for storage, virtual servers and Web services. This will set off a stultifying series of project-plan submissions, RFPs, purchase processes and a long, slow buildout of IAAS at a variety of government agencies.
Because they're the fed, most of the information about those projects is supposed to be public. Because they compete for budget dollars, someone (other agencies ) will be sure to get the word out if there's a big screwup in one of the IAAS setups.
If you never hear another word about this, the technology is stable, successful and credible.
The second thing isn't quite as boring. OpenStack, an open-source cloud-management application from Rackspace and backed by a range of other service providers as a potential standard for interaction and managmenet between clouds owned by different people, hit a major development milestone in the quest for a good, interoperableobject definition and control for cloud environments. NNASA is helping with the development, and AMD, Intel, Dell, Citrix and others are supporting it.
OpenStack is an attempt to make something exciting -- the completely undefined space between clouds and how to navigate it -- and make it predictable and dull.
So far there's not a single specification all the cloud providers and toolmakers accept or can use to make cloud-computing less interesting, so a lot of peoplel appreciate Rackspace's willingness to help develops the software and make it available as open-source.
If it, and the federal IAAS clients, succeed, cloud computing will become a lot less interesting.