Have you been following the latest tech news? Cloud has been the buzzword for sometime and now we are hearing about “This as a Service” and “That as a Service”. Services are great; it is why IT is here –to serve and protect. Before we start going into cloud services, what about the nuts and bolts of a Cloud architecture. I like to call it the FOG i.e. FOG = Infrastructure and Cloud = Apps.
According to Wikipedia, “Cloud computing is the delivery of Computing as a Service rather than a product, whereby shared resources, software, and information are provided to computers and other devices as a metered service over a network (typically the Internet).”
Furthermore, “end users access cloud based applications through a web browser or a lightweight desktop (aka Terminal) or a Mobile Application while the business software and data are stored on servers at a remote location(s). Cloud application providers strive to give the same or better service and performance as if the software programs were installed locally on end-user computers.”
So now that we have defined what we want Cloud Computing to deliver to our end users. What should the infrastructure look like? We’ve all seen those Moen sink commercials (you know the one “build the house around this faucet”), however that faucet is just a pretty silver thing unless you can deliver water to it and remove that water via a drain to a sewer. I know not a pretty thought, but think about it in terms of service delivery.
Cloud Computing works the same way, without fast and reliable bandwidth –you cannot access your applications or data –i.e. a pretty little (pretty useless) tablet in your hand. But I do have a few questions aside from the bandwidth paradigm. What about Disaster Recovery? In a Cloud world is DR good enough? I would venture to say no. Maybe Disaster Avoidance is the answer – i.e. run two or more datacenters as HOT/HOT versus the HOT/WARM scenario that we operate on today. What are your thoughts?
What about the computing infrastructure? Should we start thinking workloads versus the number of servers and operating systems (physical or virtual)? I ask this because some IT pros are anti-virtualization on the X86, maybe they are looking at virtualized Operating Systems and how they affect processor and memory usage on the physical server machine (versus application loads). My advice is to look at the applications running on the OS and define them as workloads, and then focus on the best way to utilize those resources on the physical server. We should be figuring out how to stretch those applications across multiple sites to ensure Disaster Avoidance.
We have the Technology –lets rebuild this man! Sorry Lee Majors (Six Million Dollar Man), I couldn’t help myself. But is it true we have the technology to solve the following:
- Fast and Reliable Bandwidth (Choose Multiple Carriers with Multiple Paths)
- Disaster Avoidance (Multiple Datacenters running HOT/HOT like Amazon and Google do) Maybe a play for the Cloud First initiatives from OMB
- Virtualizing for workloads versus for the 3P’s (Ping Power Pipe) stretch the loads as needed and shrink when not. (Pick your poison VMware, Hyper-V, Citrix and others).
- Bake security into the solution versus making it a bolt-on or a “one off”
I know I have covered a few topics and the purpose is to start future blog topics focused on the following: Bandwidth; Disaster Avoidance; Virtualizing Workloads; [and] Securing the whole enchilada.
So GovLoopers what are you doing in your Cloud, or should I ask, what are you doing in the FOG (infrastructure versus apps)? Do you have a secret sauce that you want to share?