Why Kubernetes Developer Ecosystem Needs A PaaS Posted on : Apr 12 - 2021

Platform as a Service (PaaS) was one of the first delivery models of the public cloud. If Infrastructure as a Service (IaaS) gave control to administrators, PaaS squarely targeted the developers through simplicity, productivity and scale.

In 2008, Google launched App Engine, a platform that enabled developers to deploy and scale Java web applications. Amazon added Elastic Beanstalk to its compute infrastructure in 2012 as a PaaS offering. Windows Azure, the initial avatar of Microsoft’s public cloud, was all about PaaS. Only in 2013, Azure got support for Linux and Windows VMs to deliver a full-blown IaaS. 

The last decade saw the rise of PaaS in the form of Cloud Foundry, Heroku, Engine Yard, and Red Hat OpenShift. 

The most significant promise of PaaS was the ability to bring source code and walking away with a URL pointing to the application. Developers never had to worry about provisioning the infrastructure, installing the OS, or configuring and securing the infrastructure. They just “pushed” the code leaving the rest to the platform.

PaaS would also scale-in and scale-out the application automatically without manual intervention. This approach freed developers from dealing with everyday operations giving them more time to focus on the code than infrastructure. PaaS led to continuous integration and continuous deployment (CI/CD), which has become the norm today. Advanced deployment techniques such as blue/green deployments, version switching, canary releases were all made possible by PaaS. View More