Is serverless the next phase of cloud-native application development?

Serverless computing is an emerging category that represents a shift in the way developers build and deliver software systems. By abstracting application infrastructure away from the code, it can greatly simplify the development process while introducing new cost and efficiency benefits. We believe serverless computing and Function-as-a-Service will play an important role in helping to define the next era of enterprise IT, along with cloud-native services and the hybrid cloud.

We have been working in the serverless space in various capacities for the better part of the last two years as different approaches emerged and communities sprang up around them. Having seen the interest surrounding these projects and evaluating their respective strengths, strategies, and visions for the future, we announced last year that we would focus our support on the Apache OpenWhisk project.

Originally developed by IBM and now part of the Apache Software Foundation, with support from additional organizations, we see the OpenWhisk project as having an advanced technology with a vibrant and growing ecosystem.

Over the past year, I have spent a great deal of time speaking with customers about what serverless could mean for them. Despite being a relatively young technology, we have seen organizations take advantage of it. In a customer survey¹ we conducted recently, we asked if their organization was currently using serverless technologies, and 36 percent of respondents to that question indicated that they are already running serverless applications in production or experimenting internally, and another 28 percent are currently planning their serverless deployments.

I have also spent time speaking with press and analysts. Across the board, these conversations have quickly moved from establishing a foundational understanding of serverless – what it is, benefits it provides, use cases where it can thrive (e.g., processing webhooks, data transformation, chat bots, etc.) – to more specific questions about putting it into practice, such as how to optimize and bring serverless into a cloud-native stack alongside existing technologies like containers. Also, not surprisingly, security questions are common. In our survey, we asked what challenges are anticipated or being experienced and 51 percent of respondents to that question indicated that security was one of their top concerns.

We've been working behind the scenes to answer questions of our own as well. In short, what does serverless mean for Red Hat?

We believe that having a serverless offering in our portfolio is essential if we want to be able to provide developers with a modern, full-featured application development stack for a hybrid cloud. We also want to stay true to the ethos of freedom. Yet, one of the biggest risks that is inherent with the traditional serverless model is that the user's applications and data can become locked into a single cloud infrastructure provider, putting their business at the mercy of whichever cloud provider they originally chose.

We believe there is a better way.

We are fortunate to be in a position of strength in the cloud application platform space with Red Hat OpenShift Container Platform, and we are building off of that strength with the introduction of OpenShift Cloud Functions, an open, capable, enterprise-grade serverless tool based on Apache OpenWhisk, that can enable customers to focus on delivering business value without having to sacrifice their freedom to do so. We believe OpenShift Cloud Functions will become the industry’s first enterprise-grade hybrid serverless offering.

By starting with OpenShift as our serverless foundation, we seek to provide portability and consistency across hybrid and multicloud environments, as well as the additional architectural components that are necessary to preserve state externally across function invocations, such as integration with an API gateway to expose functions as APIs or call external services, an in-memory datastore for faster data access, storage and file systems, user authentication, and more. These services are available both natively via the Red Hat Middleware portfolio on OpenShift, or from the cloud provider. For example, customers will be able to use AWS services with our function engine through the OpenShift service broker.

With Red Hat OpenShift Container Platform, customers can also gain powerful operational capabilities for security, orchestration, control, and scalability for their serverless applications.

We believe serverless computing is an exciting new frontier for enterprise computing and we are excited to take part in helping to bring it into the mainstream.

An early developer preview of OpenShift Cloud Functions is available through the Red Hat Developer program today, and you can expect to hear more from us in the coming months. Learn more at https://developers.redhat.com/products/cloudfunctions/overview.

¹Source: Red Hat 2018 serverless survey of 182 individuals from Red Hat customer organizations. Conducted by TechValidate, February-March 2018. Not every individual surveyed responded to each question.