How to choose the Malos Hammer Law cloud model

Technology world we usually apply the law of Malos Hammer to explain this: if you hold a hammer, all the problems are like nails. Cloud computing is no exception, everything is entering the cloud, or at least we are being told.

To conform to this metaphor, any part of the cloud that a vendor provides can become a hammer, and all of the company's workloads and applications become nails. The problem is which distributed models fit the needs of the enterprise, whether IaaS, PAAs or SaaS.

It Malos hammer law

Software as a service (SaaS) is a typical "homemade or outsourced Showdown" decision; If you can buy a service that meets most of the requirements, then this is the most economical cloud to use. But don't let the fear of building applications drive you to purchase any service that does not meet your business needs. Your business processes are a key differentiator for your business, so don't let these competitive advantages compromise your out-of-the-box choices.

If, after evaluating the out-of-the-box options, you determine that your requirements are best suited to customizing your build application, the next step is to determine if you need the cloud. Is your question a matter of scale? If not, there will be a more economical way to deal with it. If so, the cloud is the best choice, but governance or some other external agent-induced regulatory or legal issues can be a problem for you.

To infer that Izumo is the best way to customize a build application, the traditional view of the cloud vendor is to show you a choice of binary: You have to choose a platform as a service (PaaS) or infrastructure as a service (IaaS). The intention of the cloud manufacturer to recommend this option is obvious, but if we look closely, we will find that this does not hold customers.

About three scalable scenarios: transaction processing, data and geography, the IaaS layer, you can instantly extract data and geographic location, just as the storage capabilities of PAAs products are the same for Google App engine or Windows Azure or Heroku, cloudbees. In addition, geographic distribution is the basis of cloud services. Execution resources need to be chosen to conform to the requirements that are implemented as software applications.

But PAAs products, such as Windows Azure, are a valuable transaction. Developers have been freed from Low-value management functions, giving up some fine-grained environmental controls. The environment in which PAAs developers work is defined by this service. Although a product may involve changing the underlying environment, this approach is not possible in PAAs. The economic value of PAAs does not create an infrastructure picture during the involved phase, and therefore does not have to be maintained in production.

In an open mind, PAAs fully conforms to the needs of most cloud applications. However, some scenarios that operate in PAAs are too complex to make it difficult to practice. For such a scenario, IaaS is usually more appropriate. The formulation of this decision must be clear enough to have a direct economic impact. In fact, it offers opposing value transactions: Developers gain fine-grained control of the environment in exchange by accepting responsibility for building and managing the environment. This responsibility includes time to market during design time, and ongoing costs in the maintenance and operation areas.

So what is the correct rate for choosing PAAs? Apply Einstein's famous quote: You should use as many PAAs as possible, but not beyond the scope of rationality.