The transition from the server-oriented LAMP stack approach to the new Jamstack approach has proven to be a practicable concept. In recent years, a large number of Jamstack web projects have been implemented. The concept applies to smaller websites as well as complex web applications with thousands of pages. However, this conversion process should not be underestimated, because a completely new architecture is used.
The switch to Jamstack is initially associated with a degree of uncertainty. A new development paradigm is being introduced, with new rules and new processes. This conversion usually takes a number of months and initially requires operating parallel structures. However, this is nothing new for IT, and there are extensive best practices available for such a process. Additionally, as the process is better understood, the conversion speeds up drastically, saving vast amounts of time in the end.
An internal email written by Jeff Bezos1 in 2002 is legendary: all of the teams at Amazon received the binding instruction to make their data and functions available only via service interfaces in the future. Without exception. The transition to these formal APIs made life difficult for employees in the short term. However, Amazon was able to operate its systems much more efficiently and it enabled, for example, the launch of publicly available Amazon Web Services.
A key to new and secure client-side authentication is the enforcement of the OpenID Connect open standard. Technologies and identity providers such as OAuth and Auth0 enable a central single sign-on, which is now supported by all major platforms.
In addition to developing new web frameworks, the Jamstack approach requires more web components. By separating the backend and frontend, frontend architectures, browser APIs, HTML and CSS standards could evolve independently. For example, entire ecosystems are developed for ready-made APIs for authentication, for e-commerce, search and so on, or large libraries for very specialized and reusable microservices.
There were initially limitations in pre-rendering HTML markup for too many pages. Because of its size, the build and deploy cycle was very long. In the meantime, rendering tools have become so fast that thousands of pages can be processed within minutes. The advantage of cloud-based structures is also the possibility of parallelization through Lamda processes. For example, Amazon Web Services (AWS) can render thousands of pages simultaneously in up to a thousand simultaneous processes. In addition, content management systems such as Scrivito have the ability to update content after prerendering to include the latest changes to all content, so that the delivered pages are always up to date.
A geographically distributed CDN with locations close to the user delivers the ready-made content of the pages very quickly and ensures good performance. Crucially, the CDN has a sufficiently large number of distribution and end nodes with redundant locations.
A globally distributed CDN, with locations close to the user, delivers the ready-made content of the pages very quickly and ensures good performance. Native CMS cloud solutions like Scrivito have already integrated the CDN requirements conceptually.
The Jamstack approach has several key advantages over the traditional server infrastructure. It enables higher performance, improves security, lowers costs, scales more easily, and delivers a better user experience.