A cloud infrastructure can use many different technologies during normal operation. In this video, you’ll learn about internal vs. external clouds, metered cloud usage, elasticity, availability, file synchronization, and multitenancy.
Many organizations will build their own private cloud. They have their own data center that they manage themselves. They will purchase their own equipment and their own storage, and they will install and maintain everything themselves. Sometimes you’ll see this referred to as an internal cloud or a private cloud.
All of those resources in their private cloud are dedicated for their use only. No one else will be using any of the resources that they have acquired and that they maintain in their own internal cloud infrastructure. There is obviously a significant cache outlay to be able to purchase all of this equipment initially and install it into your data center, but once everything’s in place, you no longer have those pay-as-you-go or metered costs associated with a cloud infrastructure. You can use as much of your private cloud as you would like, and there’s no additional cloud costs associated with that.
Some organizations prefer not owning any equipment or maintaining any data center, and instead will use a cloud that was created by others. In this case, it would be an external cloud or what we often refer to as a public cloud. So if you are using public cloud services on Amazon, on RackSpace, on Microsoft, then all of those cloud services would be considered a public cloud.
There are many other customers that are deploying their own cloud-based services, so both you and those other customers are all sharing the same resources to be able to deploy all of those cloud-based technologies. Obviously, in these public clouds, one customer is not able to see any data from another customer, but all of you could be using the same storage drives, the same CPUs, and the same network as every other customer in that public cloud.
And although you didn’t have to pay for any additional costs up front to put together this cloud, you still have to pay for the resources that you’re using. Some cloud providers charge you a fixed rate per month, or they might charge as you go in more of a metered pricing. So your organization will need to determine if a private cloud might be something worth investing in, or if a public cloud might be a better option.
For many organizations, costs are certainly an important consideration. And with cloud services, you could be billed differently depending on the cloud provider and the service that you’re using. You may find that some of your cloud services are billed back to you using metered utilization. This means that you as the customer are paying for how much of these resources you are using.
So if you’re paying to have traffic uploaded to you inside of this cloud, all of that traffic would be ingress traffic, and there may be a cost associated with every bite that you would transfer of that ingress traffic. There’s probably also a cost to store the data on the storage drives of that cloud-based service. And if customers are downloading information from your cloud services, then you are sending that information to the user and all of that information would be priced on the egress traffic costs.
And if customers are downloading information from your cloud services, then you are sending all of that traffic out. Or it would be egress traffic, and there might be a different cost for egress traffic versus ingress traffic.
At the end of the month, all of these costs are added up, and they provide you with an invoice of what you would need to pay for that month. If you have a very busy month, then you probably are paying more than if it’s a slower month because it is a metered utilization or based on how much you’re using.
On the consumer side, we tend to see a lot of cloud-based services priced on a nonmetered basis. This means that it is a fixed price for a certain amount of resources. And those services generally charge you based on the maximum amount of storage that you could use.
For example, you might pay a certain amount for two gigabytes of storage in the cloud. And no matter how much you upload or how much you download, you’re always paying the same amount every month for that maximum block of storage.
One of the most useful characteristics of cloud computing is the ability to scale up or scale down at any time. This means as customers or applications become more active, you can scale up your infrastructure to be able to handle that load. And perhaps at different times of the year when things get a little bit slower, you can scale down your infrastructure so you’re only paying for what you need.
And of course, you’re able to make these changes at any time, and very often automatically. So it would not be unusual for you to push a button and instantly deploy a new application instance to be able to handle the additional load. We refer to this ability to instantly scale up or scale down based on the load as elasticity.
Another very appealing characteristic of cloud technologies is that it always seems to be available. This ability to maintain uptime and availability is an important characteristic for any infrastructure, but with cloud-based technologies, there are many redundancies in place across many different data centers.
This means if one data center is having problems with a particular drive array or with a network connection, that other parts of that infrastructure can pick up the load. And in very extreme cases, some of these cloud providers will have completely separate data centers, and they can move everyone from one data center to the other to maintain that uptime.
The cloud also allows us to appear to be everywhere all at once. This means that we can have a database that we’re maintaining here in North America and be able to have that information automatically duplicated to other data centers around the world. Sometimes this file synchronization is something built into the application and is maintained by the customer, but some cloud infrastructures will automatically synchronize this data behind the scenes, especially if they need to move you from one data center to another.
And to be able to get the best efficiency for the infrastructure that’s been installed in these cloud-based data centers, we need to have many customers using all of this equipment all the time. We refer to this characteristic as multi-tenancy. That means there are many different tenants effectively living in that cloud infrastructure, and they are all working and operating simultaneously across all of those resources.
There are obviously technologies, processes, and procedures in place to maintain separation between all of these different customers. But by having all of these customers use all of these cloud-based services simultaneously, the cloud provider can provide an efficiency in technology and an efficiency in cost.