In the last decade, a number of cloud providers have come into existence to help provide support in regards application deployment. Each of these cloud providers offers a distinguished set of features to make their service stand out from the competition when it comes to luring the organizations and developers to use their platform for the deployment of their applications.

The shift toward the cloud deployment model provides various kinds of advantages for the developers/organizations that are responsible for the development of the applications, including the following:

• Reduced cost of infrastructure maintenance: As the application deployments move to the cloud, the costs of maintaining the infrastructure are reducing. This is happening due to the fact that the cloud providers are now responsible for maintaining the hardware on which the applications are running, and the individual developers and organizations do not need to purchase this hardware and handle any kind of issues that may happen with it.
• High uptimes: Most of the cloud providers guarantee a high uptime for their infrastructure, which is made possible due to the high amount of infrastructure replication done on their end. The end beneficiaries of this are the developers who are maintaining a particular application in the cloud, because now they can provide a high uptime to their application users without worrying about the productivity losses that may happen if the infrastructure goes down.
• Low latency: Inside the cloud deployment approach, the developers can aim to provide low latency to the application for the users. This is made possible by replicating the application instances across the different geographical data centers of the cloud service provider. Once the application is replicated, the cloud service provider then re-routes the requests to the application servers, which are in close proximity of the client so as to facilitate low latency responses.
• Easy scaling: As the load on the application increases, new application instances might need to be spawned up to handle the increased load. The cloud service providers usually provide the facility of dynamically scaling up the application as the load on it increases and scaling down of the instances as the load subsides. This provides a high throughput, low cost solution to handle the peak loads without worrying about the manual intervention that is usually required in the traditional infrastructure. Also, the response times associated with this scaling are usually low in comparison with what is present in the case of traditional infrastructures.

All of the preceding points make a compelling argument for the shift to cloud-based deployments for the applications. But based on the needs of an organization, they may or may not want their application to be deployed on a third-party server where the organization barely has any control. To handle such scenarios, the organizations may decide to move toward their own private clouds that run on their infrastructure and handle the deployments of all the applications inside the organization. So, let's spend some time understanding the various models of cloud deployments that are out there.