The computing paradigm commonly known as "serverless" computing isn't for everyone, but it does have a place and plenty of opportunities for those willing to explore it. IBM has its own serverless platform called OpenWhisk, which first became generally available in December 2016.
In a video interview with ServerWatch, Jason McGee, VP and CTO for IBM Cloud platform, discusses the opportunities for serverless, event-driven computing and where the technology intersects with Watson cognitive computing and the application container revolutions.
Among the interesting real-world deployments that IBM has for OpenWhisk today is with a European bank for check processing. The system enables users to take a picture of a paper check and then execute a bank deposit online, all done via the OpenWhisk serverless back-end infrastructure.
Serverless vs. Auto-Scaling
The idea of being able to scale up compute resource on demand is not a new one, and in past years it was typically handled with server virtual machine auto-scaling. McGee explained that serverless and auto-scaling are in fact two different things.
"With the check processing example, it's actually a discrete event, and that can help drive the scaling that is different than traditional auto-scaling that determines when the CPU is busy," McGee explained. "For event-driven workloads, the serverless model is much more natural from a runtime model and also from a developer model."
"With serverless the term is actually unfortunate," he added. "The value to me is less about it being serverless and more about the event-oriented nature."
Watch the full video interview with McGee below:
Sean Michael Kerner is a senior editor at ServerWatch and InternetNews.com. Follow him on Twitter @TechJournalist.