A private cloud computing platform is a stack of network, server, and storage hardware dedicated to you for the purpose of cloud computing. When you use a managed cloud computing service, your hardware stack becomes a customizable cloud of compute and storage resources that can be configured and reconfigured whenever and however you want. Why is this ability to configure and reconfigure your server resources with a private cloud computing platform so valuable? This is why.
With a typical dedicated server stack, which you manage or outsource, you select your server, storage, and networking needs, buy them, and then live with that configuration for 3-5 years. Over the course of those years, you can add memory, which isn’t too difficult. You may need to upgrade disk drives, which is a bit more difficult. Over time you may need to upgrade the CPU, which is very difficult and expensive because it basically requires changing the entire server.
When talking to financial professionals, the following analogy can be used to describe the benefit of private cloud computing:
Imagine that you were responsible for bus transportation in a metropolitan area. He is preparing to order new buses that his community will need to live with for the next 7 years. The buses are available in 20, 30 or 50 seat configurations. You need between 1,000 and 1,100 seats in total to accommodate your population and the routes you have designed to meet their needs. Each line is designed to maximize the use of a 20 or 50 seat bus. So you take a guess and buy twenty 20-seat buses and twelve 50-seat buses, and then expect it to be efficient for the next 7 years.
Interesting puzzle, huh? Even if it’s done really well, at the end of the day, there’s still plenty of room for inefficiencies and wasted resources.
Instead of being forced to live with the same configuration of resources for several years, imagine that you can buy a “cloud with 1,000 bus seats” and reconfigure those seats at will into any size bus you want, at any time. The buses can be as small as single seat or you can have a bus with 1000 seats. Do you have a convention? Make 4 x 200 seat buses between the hotels and the conference center. Having this flexibility is really powerful and valuable. Why? Because with the increased flexibility, you can have more seats in use at any one time, requiring significantly fewer seats and providing a better experience.
That could be called “cloud transportation,” and it would be every transportation official’s dream. Now think of each bus as a server and you have “private cloud computing” – every IT CFO’s dream come true.
An important new requirement
We have just seen the power of flexibility provided by private cloud computing. Unfortunately, it introduces a level of complexity that was not present before. There are some unique skills that need to be in place to really get the benefit of a private cloud.
Continuing the transportation analogy… now you have this new on-demand flexibility for your buses. All you need now is a team of people to control the flow of traffic and redesign buses with more or fewer seats based on demand and usage. They would work feverishly to reduce waiting and route times for anyone who wanted a seat on any bus. Flexibility is powerful, but to take full advantage of it, you need quite sophisticated and complex (ie expensive) skills to dynamically change your settings and reap the full benefits of flexibility.
In the computing world, the thing to do is keep an eye on server, storage, and network resources. When one appears to be causing a bottleneck, provide more resources. You should also remember to reduce the amount of resources allocated to a server that doesn’t need it. If you don’t, it remains inactive and unavailable to another server that might demand it. This resource allocation process is powerful, but not cheap.
That is where you should pay attention when designing your Private Cloud. You need it to automatically change cloud settings in real time so resources are where they need to be when they need to be there! In the blink of an eye and automatically, it should turn off cloud servers that are not being used and turn them back on when they are needed. This means that a small, extremely smart piece of software will have to constantly monitor your server, storage, memory, and network resources and compare them to workloads. It will then estimate and forecast which servers need the most resources. After estimation, it should automatically reallocate resources in real time so that you are always using your cloud computing resources in the most efficient way. How cool is that? The dream continues…
cost
The beautiful ending to this story is that the final savings from idle capacity can be passed on to you. The cost of a well-designed private cloud computing platform is less than a dedicated server per server! So not only is it more flexible and can offer a lower total cost of ownership, but a managed private cloud can be absolutely cheaper. That is the benefit of private cloud computing.
