Cloud computing over Client-server: differences, cons and pros?

In the great-wide world of computing, there are only two kinds of applications, standalone application, and client-server application

Cloud computing over Client-server: differences, cons and pros?


In the great-wide world of computing, there are only two kinds of applications,

  1. standalone application, and
  2. client-server application

Standalone simply means the user-interface and the business logic and the datastore and any and all resources are completely self contained within a single execution environment - the end-user's machine.

In contrast, client-server means there is some logical segregation of these components, services, and resources. Typically we see user-interfaces in a local execution [aka "client"] and business and data store and other resources in a remote execution [aka "server"]. Some examples,

  • MSN Messenger [contains an application we run on our local machine, while a back-end service facilitates communication at a remote location]
  • Facebook [contains a thin application - our web browser - and a remote service - their servers]
  • StackOverflow [same Facebook]

With this in hand, let's look at cloud computing.

Cloud Computing

Cloud computing is an abstraction of traditional server hosting solutions. Instead of buying 10 servers myself to run and manage in my own operations datacentre, I now lease X servers from a vendor where X is a variable number decided by me whenever I want.

There is a distinct advantage to leveraging a cloud. If I bought 10 servers, I must manage and maintain these 10 servers even if they are underutilized [say only 1 server is used 90% of the time, while all 10 are pinned 10% of the time at peak hours]. That means I am paying way too much in maintenance for 90% of the time, while being inflexible when I need to grow the remaining 10%.

The advantage of cloud computing is that "someone else" is managing the server farm for us, and is willing to lease out a variable number of machines to us on demand. So in our scenario above, I could lease 1 machine for 90% of the time in off-hours, and scale up to 10 or more machines the remaining 10% of the time.

Microsoft takes this abstraction one step further with Windows Azure. They do not lease generalized servers, but application domains. This is the one example [to mind] where a cloud implementation has actual design implications - but it is predicated on the same premise of elastic hosting.


For the most part, because client-server is a software paradigm and cloud computing is a hosting abstraction, they are independent concepts. Keeners, however will realize that cloud computing implies business, data stores, and other resources are remotely hosted, which necessitates that any application running within a "cloud" is part of a client-server application.