Arrived at the session early and was easy to get in even though I hadn't registered. The CTO of Stratus Tech started and covered the basics for his presentation. Covering things like his history and what the cloud is. A pretty common "Cloud is whatever folks want" and 80% of IT management agrees that the cloud will be great. He feels that SaaS will have a .com style consolidation any day now.

Much of the legacy technology architectures like Mainframe, client server and n-tier applications are not going away even though everyone claims they are dead. As such the cloud is just moving these legacy systems into a separately managed group. Not necessarily removing them from service.

One of his big points is that Availability is a bit concerning moving to the cloud. Application Platform, Vendor Trust, Mgmt/Platform monitoring & Billing are 4 major areas of concern to make sure you know and have defined in the contracts. Can you trust your business to these when the services are down? VMware offers some of these.

The Stratus CTO's comments are that you need a cloud environment with Fault-tolerant hardware platforms, 24/7 services for support, Culture of high focus on availability technologies, and 30 years of experience doing work like this. 99.999% hardware uptime accomplished by heavy investments into monitoring and operational simplicity.

Stratus provides and develops lockstep hardware technology to keep things running as they provide hardware tandem architectures. 8 of 10 banks, 10 of 13 pharma, 900 health agencies. This is clock locked systems in sync. MasterCard & Visa use this so things always work. 100% of credit card transactions in Japan are run through their servers. They work on Windows, Linux and now VMware.

This kind of hardware takes you from Basic HA to Better Fault Tolerance to finally Continous Availability. The FAA has identified that Continous Availability is extremely critical. They are looking to upgrade from older versions of the hardware and Stratus's answer was to move to vSphere with Stratus's ftServer to get to 24 hour and near 100% uptime.

Virtualization works today for much of the functionality of "cloud" In the near term their will be Private Clouds. Moving to public clouds is still in the air and there is a lot of hype versus reality. Security and availability is a major concern for public clouds.

What keeps clouds up? Products, Services, Experience & People and Culture. It is the whole picture and not just pieces and parts. What is the cost of downtime and is the possibilty of public downtime acceptable? Many pieces of technology are extremely difficult to deal with in this regard to validate uptime and keep your business working.