Whether it be proprietary options such as WebSphere or WebLogic, or open source alternatives JBoss or Tomcat, all are in decline based on employment listings we review. However, they are not declining at the same pace. From our collection of data, WebSphere and WebLogic's decline has been more muted. The rate of reduction for each of these application servers is in the neighborhood of 25-35% over the last couple years. At the same time, the likes of JBoss and Tomcat have declined around 40-45%. Not a drastic difference, but one that still is notable.
Why are the FLOSS based application servers losing ground at a faster clip than their proprietary brethren? I am sure there is a multitude of possibilities, but one that might glean some insight is the progressive nature of different companies. It is more likely a company that relies on proprietary solutions is an established entity that moves at a slower pace when considering changes. There is a strong probability it will keep the existing application server around for a longer period of time before switching gears. Meanwhile, companies that have FLOSS based solutions are more likely to be more nimble. They probably have a tendency to keep up with the trends more aggressively, and as a result do not hesitate to make a change when there is a technological advantage in doing so.
Outside the business prognostication, there are some seismic transitions that have occurred within engineering departments. Primarily the processes that they follow. During this time of application server regression, we have witnessed the rising of DevOps, Micro Services, Serverless and Continuous Delivery. Each of these have had a profound effect. Whether it be an application being developed using Continuous Delivery resulting in the need for multiple deployments daily, or Micro Services' need for a taxing multiple application servers to run each component, the servers bog down the potential efficiency. On top of that, the popularity of structuring a department in a DevOps fashion has limited the need for an application server. Ultimately, the amount of employment listings proclaiming a need for proficiency in these three categories has jumped significantly in the last five years. Hence, it is hard to dispute their importance in this trend.
Besides what has been mentioned, there are a multitude of other factors that might be up for debate. However, one can not ignore the current decline of application servers no matter what there makeup is. It appears more and more entities are realizing they do not need the complexity and lack of efficiency of utilizing an application server to deploy applications. Engineering departments are in a constant state of reconfiguring their processes, and in the case of application servers, it appears that the part they once played in the overall infrastructure is no longer a necessity.