The new Fission Workflows product sits on top of Fission, which itself sits on top of Kubernetes as its orchestrator. Fission Workflows allows developers to map out sequences of serverless functions to garner more capacity and scalability out of those deployments.
Soam Vasani, software engineer at Platform9 and Fission author, said the Workflows product uses automation to make it easier for developers to arrange complex applications. These applications can then run in private data centers or on public clouds. Access is provided through a centralized dashboard that offers a view into the workflow’s execution.
The product looks to offer what many serverless and container platforms promise: remove the developer that is writing application code from having to know – or care – where that app will run. It also allows those developers to re-use some code components to reduce overall workload.
“With Fission we saw a lot of traction, including some use cases, bots, and web hooks,” Vasani said. “But there is the question of what to do with more complex applications.”
Soam described Workflows as basically a flow chart for how to run different functions. “It’s really useful for automating the interaction between functions without writing complex code.”
As serverless computing advantages are often tied to the technology’s agility, Soam said Platform9 focused on removing operational overhead from the Workflows platform.
“The function runs the same way as it would without Workflows, which operates on the control plane,” Soam described. “The function still runs in a container on Kubernetes. We do not add any additional overhead to the actual function.”
Serverless computing by its nature is designed to reduce the amount of overhead associated with offering services in the cloud. This includes the ability for a cloud provider to dynamically manage server resources.
Soam also explained that Workflows can speed up serverless functions and reduce operational costs by “preloading” a function. In reducing potential waste and taking advantage of their diminutive resource needs, serverless computing allows for pricing models that typically break down access by tenths of a second.
“Because Workflows knows a function is about to run, it can preload or prefetch that function,” Soam said. “This can allow for scaling up a function in advance of a request coming in, but also remove the need for any unnecessary scaling.”
While Workflows is now available, Soam said Platform9 expects to roll out new functions on a monthly basis. This could include the ability to integrate workflow control over serverless functions running on other platforms like Amazon Web Services (AWS) and Microsoft Azure.
“These are all things that we are looking at, but for now we are focused on making it work well with our Fission product,” Soam said.
Platform9 in June raised $22 million in a Series C funding round, which pushed its total VC haul to $36.5 million.