To get OpenStack Cloud, get this: developers want Serverless Computing
Just back from a pretty intense week at the SRO OpenStack summit in San Diego (capacity: 1200; attended: 1500), and I can report that the hype about the hype is not overhyped. There's an early feeding frenzy on, as companies at various points in the supply chain jump in with both feet to figure out what it's about and what to do about it. (We did our part to answer that question on the commercial side; see Boris' blog post from Wednesday).
The good news is that a lot of the folks I met from Enterprise Technology Incumbents (ETIs; you know who you are) told me they attended because their big customers told them they wanted to know about OpenStack, and want to get going on it with vendors whom they trust. Many of these vendors successfully got on the boat with Linux (could be big!), and want to know what's next.
The bad news is, particularly for some of the storage and networking ETIs, based on some of the questions I heard from the _technical_ people, is that they have not shifted their thinking to the world where The Developer is The New Kingmaker. (Memo to me: get Redmonk's Stephen O'Grady, whom I had the pleasure of seeing at the summit, to future OpenStack Summits to talk about who doesn't get this and why).
A very useful addition to the discussion that I would recommend strongly to all ETI technologists (unless they're counting on using their COBOL skills for the next Y2K) is the notion of Serverless Computing, a great post from Ken Fromm, a bizdev guy at Iron.IO:
THINK SERVERLESS: The phrase “serverless” doesn’t mean servers are no longer involved. It simply means that developers no longer have to think that much about them. Computing resources get used as services without having to manage around physical capacities or limits. Service providers increasingly take on the responsibility of managing servers, data stores and other infrastructure resources. Developers could set up their own open source solutions, but that means they have to manage the servers and the queues and the loads.
Multiply this effort by the number services an app might consume (task processing, message queues, SMTP servers, payment services), hosted services quickly start to look like the future of computing.
To this I would add, consider that developers won't care about IP address either. Our customer Reinhardt Quelle of Cisco Webex, put it something like this in answer a question from one networking infrastructure engineer: "I care about security zones, not the port policies." vLANs, ports, all that stuff is still important -- just like tracking memory addresses or poking registers -- just not all the time or as a condition for getting anything done. Look for Storage as a Service and Software Defined Storage to help drive these changes to those resources, too.
Interestingly, even HPC is moving in that direction; it's not a particular server, but a pool of resources, well configured and orchestrated, that make it possible to get work done; Australia's 4000-core research cloud, NeCTAR (with the help of our friends at Aptira among others) is a great example. Nebula's Chris Kemp showed an interesting diagram in one of his characterically fine speeches (I'll link it once he posts) that suggests that the 'stack' is best shown as a set of concentric circles and segments, representing services, with applications in the middle.
It's a heady time for OpenStack, and things are picking up speed. To achieve escape velocity, we need two things, I'd say:
(1) the hype crescendo: it's working. And as Abraham Lincoln observed: "“With public sentiment, nothing can fail; without it nothing can succeed.” (Citation credit to Redmonk's O'Grady, again)
(2) The proposition will work for companies who need new, lower cost, higher productivity application development and execution. If OpenStack focuses on making it easier for developers, they'll pair up with the CFOs looking to drive down cost, and the flywheel will be in motion.