Free image by Gerd Altmann from Pixabay

A shorter version of this post originally appeared on ibm.com/blogs

If you have been around the software industry long enough, you have seen architectural changes and new computing paradigms from time-to-time.  The history of computing is well documented, and it continues to evolve. From monolithic to client-server architectures, and from web applications to containerized cloud-native applications, we see a constant evolution with more software available thanks to the proliferation of open-source software.

Over the last two decades, open-source software has grown exponentially. Having software publicly available and the opportunity to collaborate, contribute innovations, and fix bugs has become the building blocks for creating more open-source and commercial software. There are many great examples of software created with open-source, as well as open-source enabling the creation of more open-source.

Driven by open-source technologies, serverless computing is a computing paradigm that has been around for a few years, and all public cloud vendors, including IBM Cloud, offer the service.  Serverless computing typically refers to the approach of building and running applications hosted by a third-party, but unlike cloud computing, you do not manage or even access the infrastructure.

The hosting of serverless applications is only part of this new computing paradigm. The most important aspect is the model of breaking up applications into individual functions that can be individually invoked and scaled. This more finely-grained development and deployment model allows for applications to have one or many functions that can be executed and scaled up or down on demand.

There are plenty of documentation and marketing materials addressing the benefits of serverless computing, most of them focused on the benefits of not managing servers and only paying for the functions executed with no cost for idle capacity in virtual machines and containers.  The focus of this blog goes beyond these topics; it is about the opportunity to bring a new architecture and the evolution needed to build and deploy software for highly transactional applications that will benefit greatly from an architecture based on functions and events.

Today we are announcing that you can now develop and deploy serverless computing on IBM Z and LinuxONE with Red Hat OpenShift Serverless.  Well-aligned with IBM’s hybrid cloud strategy, this new offering is available as a no-charge add-on to the Red Hat OpenShift Container Platform.

Let’s first explain how Red Hat OpenShift Serverless relates to other cloud-native initiatives.  Serverless computing complements other cloud-native applications. In fact, the Cloud Native Computing Foundation (CNCF), which helps build sustainable ecosystems for cloud-native software, has been a key promoter of serverless computing, including driving standards such as CloudEvents, a specification to ease event declaration and delivery across services and platforms. CNCF hosts important open source projects that are driving the growth and progression of serverless computing. 

A key component of cloud-native applications is the use of containers, and orchestration for containers is done by another important open-source project, Kubernetes. 

One of the most complete open-source projects for serverless computing is Knative, a platform that provides components to build and run serverless container-based applications on Kubernetes.  Yes, you can create container-based applications broken into one or multiple functions to run those functions on stateless containers.

Beyond auto-scaling for HTTP requests, you can trigger serverless containers from a variety of event sources and receive events (i.e. GitHub and GitLab events) and database messages.

Red Hat OpenShift Serverless is based on the Knative project. Applications are created using Kubernetes custom resource definitions (CRDs) that define and control the behavior of serverless workloads on an OpenShift cluster.

Red Hat OpenShift Serverless capabilities include Knative Serving, Knative Eventing, and Knative CLI. You can deploy serverless applications in practically any programming language and enable auto-scaling to scale up to meet demand or scale down to zero when is not in use.  You can invoke serverless functions using plain HTTP requests or CloudEvents. You can also trigger serverless containers from a growing number of event sources, and Red Hat OpenShift Serverless comes with out-of-the-box project templates to jumpstart your code.

As is customary, software developers from IBM and Red Hat contribute to open-source projects including all Knative components, and to make sure we continue to enhance functionality and ensure it works smoothly in the s390x architecture for IBM Z and LinuxONE.

This is a significant new capability for all those Linux on IBM Z and LinuxONE deployments in the largest and most important enterprises in the world.  You now have the option to migrate transaction-intensive applications or to build new applications based on individual functions while combining both containers and serverless functionality.  Open-source software, and specifically Red Hat OpenShift Serverless for IBM Z and LinuxONE, gives you the framework to use containers and also serverless functions. Actually, developers can use a single platform for hosting their microservices, legacy, and serverless applications.

Moreover, to round it all up, applications can be part of DevOps automation. The ecosystem of open-source software on Kubernetes continues to grow and one example is what originally started as a Knative component and then branched out to become Tekton Pipelines. Just like Red Hat OpenShift Serverless, it has become a no-charge add-on offering for OpenShift known as Red Hat OpenShift Pipelines also available on IBM Z and LinuxONE.

Now that you have the opportunity to move into serverless computing, it could be ideal for workloads that require parallel processes. Serverless computing is well-suited for event-driven, stream, and message queues. Most applications with large volumes of transactions, including AI-related ones such as Monte Carlo simulations, database updates, and event processing of data with small payloads, (i.e. IoT) are ideal for serverless computing. Moreover, unlike Function-as-a-Service offerings, with your IBM Z or LinuxONE, you do not trade control and visibility of your infrastructure when you are doing “serverless.”

To summarize, you now have the opportunity to develop applications based on functions with discrete units of code for event-based execution. This will bring development velocity and rapid benefits to the business.  You are going to be able to take advantage of the underlying scalability and flexibility of IBM Z and LinuxONE.

Try a new computing paradigm on IBM Z and LinuxONE with Red Hat OpenShift Serverless and stay tuned for more OpenShift add-ons, support for more event types, and more great technology powered by open-source innovation available to you in integrated and easy-to-deploy supported packages.

For more technical information and a few screenshots read this other blog post.