Clouds are not “clouds” because of technical reasons. Cloulds are “clouds” because of business reasons. What’s ec2? Its Linux, its Xen, its some Ruby-based tools, its a billing system, and its a whole lot of commodity hardware and network gear. There’s really no new technology there. What actually is new? How it’s being run, the business relationship between you and the provider, the financial model under which you get to use the stuff, the ownership of the operational risk… these are all business concepts and this is where the innovation lies.

The IT management stack (hardware and network provisioning, os provisioning, platform configuration management, application deployment, monitoring) is the same inside the cloud as it is outside of the cloud. The only differences are who is responsible for doing it and the level of automation and reliability that their business needs require.

In my own self-interest, the whole cloud concept has been great because it shins a light on just how semi-manual, kludgy, unreliable, and downright awful most people’s processes really are. But I hold out little hope for this new buzz to create any sort of IT management standard. It hasn’t happened since the invention of commercial computing and it’s not gong to happen now. All we can do is focus on making straightforward and effective free tools that solve peoples problems and easily interoperate.

Keep up the good work, John.

-Damon
http://dev2ops.org
http://controltier.com

I think the magic of the “cloud” is going to come down to cost and speed of deployment. Cost savings will come in a number of ways. One, the obvious savings for new projects, startups, and prototypes. I spoke to a startup the other day that is using EC2/S3 and they estimated that they saved at least 70k for their initial startup. That is huge for a few guys trying to take a new idea to market. The second big saving is when you get into template based clouds like 3Tera, RightScale, and Elastra the administration costs can be significantly reduced. These type of infrastructures start to save companies huge administration and professional service costs because the complex multiple instance configurations are simplified by tools and scripts. For example, the 70k savings that the startup I mentioned was only in hardware infrastructure costs. They still had to design,develop, and implement a complicated logical multiple instance infrastructure. Fortunately , for them, they were an ex Java professional services guys but their time/costs could have also been reduced had they been able to start with someone like Rightscale and Elastra.

cloudstatus.com showed the basics of what is possible. When cloud providers start seeing their resources graphed, in public view…now we’re talking competition.

If I didn’t know who you were I would be an idiot. I consider you one of the rock stars of this movement. I am really sad I couldn’t make it to Velocity 08… hopefully next year.

Thanks
John