Getting Started with cloud computing: How do you evaluate cloud computing performance?

You need to consider several performance patterns; performance patterns can be grouped into three very basic categories:

• Customer-facing (lack of performance)

• Cloud-oriented (performance advantage)

• Mixed type (performance is difficult to say, depends on the implementation)

Customer-facing cloud architectures refer to these systems: cloud computing Providers (typically SaaS providers) interact with users over the Internet. The problem here is not that cloud providers are slow to deliver services, but that the machines that are constantly appearing between SaaS providers and browsers have a delay in the machine's back-end session.

There is little you can do about it, apart from creating an expensive proprietary link between your company and the SaaS provider, but doing so will immediately reduce the value of SaaS. Customer-facing platforms are obviously not as fast as the applications running on the local network, but in most cases the user will not notice the delay unless there is network saturation.

Cloud-oriented cloud architecture refers to these systems: processes are carried out within the cloud. Most infrastructure, namely, service (IaaS) providers and some platform-service (PaaS) providers fall into this category. Such systems often provide performance over internal deployment systems because they have access to much more virtualized resources and are able to allocate these resources dynamically.

For example, you may have a database query that causes internal internal systems to saturate and requires multiple hours to run. By contrast, how many servers are needed to support a dramatically increased amount of processing, and how many servers are available for a cloud provider with sufficient resources to meet the requirements, and the result set can be returned within minutes. This reflects the flexibility and scalability of cloud computing and is a significant selling point for cloud computing when it comes to balancing performance and cost.

As you may have guessed, the hybrid cloud computing model takes full advantage of the customer-facing cloud computing model and cloud-oriented cloud computing model, often used for platform-service systems, but the two cloud computing models share a different proportion. The idea is that you can mix user interactions with highly scalable, on-demand backend processing. The downside is that the more frequently you connect to the browser, the longer the delay you bring to the pattern. Thus, the solution to this architecture is to optimize the interaction between the cloud-based back-end server and the browser. You can do this with some old-fashioned distributed application architectures.

If you're migrating to cloud computing, don't forget to consider its performance as good, bad, and medium.