Is it time for cloud standards?

Ben, 2 thoughts. As I posted on Twitter, it's hard to have cloud standards when we're still not at a point where we have common definitions.

But it's probably "easier" (relative term) to have standards at the infrastructure level than at the application level...which is where most end customers care about. Agree? Do you think one layer is more critical than another?

@Glenn - agreed it's probably easier at a lower, infrastructural level.

vmware & DMTF President Winston Bumpus and I discussed this at length (in the context of OVF initially) and agreed that “to drive true cloud independence and to become a true commodity, it has to become as standardized and available as USB”. (That’s interesting in and of itself as once USB was de-facto, a lot of companies made a lot of money on selling products based on it using the embrace and extend concept - so if vendors are worried about being proprietary to make money, maybe that is backwards thinking ?). Does standardized mean open standards ? Not always, but in this case, I think it is an absolute requirement.

There are tools today that allow me to build, deploy and manage services based on cloud, but I don’t have a great degree of confidence that I can seamlessly move these across providers with 100% success. Notice I have never said “instantly”. I am willing to be “real time enough” to ensure that I get my service up and running somewhere else, exactly as I left it before.

Is it time now ? Probably not. But it will certainly continue to be a source of great debate for the forseeable future.

Standards are okay as long as they do not stifle advances. As a utility, their need to be robust agreements moving forward. I would offer my services to help chair this group.

Christian,

I would not necessarily agree that standards, portability and independence = commodity. If that's the end goal, I don't think you will find much support for standards development. Further I don't think that we should conflate standards and portability with commodity. Perhaps at least trying to adopt something along the lines of the SaaS Bill of Rights that Ray Wang put together is a better place to start than defining technical standards. The standards need to emerge from the outcomes the industry wants to achieve. Agree?

I agree with Glenn and Neville. We need to keep the momentum and again not choke off or stifle advances in this technology.

I looks like unanimous consensus here.

While formal standards are excellent tools for interoperability (TCP, HTML, XML come to mind first), the standards-by-committee approach is not right at this time--for two reasons:

1. Cloud computing is still in the formative stages. Standards could "lock in" early non-optimal solutions, stifling innovation
2. "Cloud Computing" is still used too often as a marketing term. Even though NIST has some definitions in place, use of the term is still fuzzy. (Check out the thread on "What is your Definition of a Cloud" started yesterday for more)

We likely need to let the definition of a true cloud firm up more, then allow innovation to bring winning ideas to the forefront before formalising standards.

That being said, it would not help organisations who are considering clouds to draw up some evaluation criteria around items like the following help them pick solutions that best meet their needs: self-service, access, measured service (availability, performance), elasticity (both speed of response to changes in demand and performance as usage 'stretches'). Ben and others also point out the importance of portability. This will be even more important as the leaders change places in the market.

Clouds computing is much too nascent an industry for for

Thanks Ben for starting an interesting discussion- sorry for jumping in late here!

In terms of standards- I could come across heavy handed- but consistently when talking to prospects and backed up by quantitative research, buyers are put off Cloud because of Security and Data concerns.

And the reality is they have reason to be concerned! Standards around data security in the Cloud are far and few between, and there is little or no guarantee if the Cloud provider is doing the right thing. You'll only know when there is a breach.

Compare that to the tried and tested methods of keeping your data behind a firewall where you can prod and back it up.

It's all about control and understanding of controls. The sooner we have standards where these make security of providers much more obvious, the easier it will be for enterprises to adopt Cloud. We take security really seriously, and invest a lot of time and effort into it. The sad reality is that lots don't- and customers will get burnt.

The National Institute of Standards and Technology (NIST) have created 5 Working Groups to address current issues with Cloud Computing and other light technologies. The working groups are Reference Architecture and Taxonomy, Standards Acceleration to Jumpstart the Adoption of Cloud Computing (SAJACC), Cloud Security, Standards Roadmap, and Business Use Cases.

The charter of the Standards Roadmap group is to survey the existing standards landscape for security, portability, and interoperability standards/models/studies/etc. relevant to cloud computing, determine standards gaps, and identify standardization priorities. The primary deliverable will be a recommended Cloud Computing Standards Roadmap document, and supporting deliverables will be developed as necessary. Anyone can join the working group at http://collaborate.nist.gov/twiki-cloud-computing/bin/view/CloudComputing/Web... and review the ongoing work and materials. We expect the first formal draft of the roadmap in March, and release of the draft for comment at the next NIST Cloud Computing Symposium April 07-08.

If you want to be part of the U.S. national discussion about how we are going to achieve Cloud Interoperability and Data Portability, the Working Group would welcome your participation and input.

At this time, cloud standards will be more counterproductive than helpful. Standards will help the larger and slower players with a lot of inertia hold back the pace of innovation in the sector. We have just delivered the fastest desktop ever with NitroDesktops, and we have just begun. Lets hold off standards for now.