OpenFaaS on Kubernetes has scaling to zero turned off by default, which means there is no cold start. It can be turned back on for various functions where latency isn't important or where they are run asynchronously. On a single VM with the faasd distribution scaling from zero is extremely rapid.
Thank you for mentioning openfaas.
OpenFaaS on Kubernetes has scaling to zero turned off by default, which means there is no cold start. It can be turned back on for various functions where latency isn't important or where they are run asynchronously. On a single VM with the faasd distribution scaling from zero is extremely rapid.
Check out the video presentation that explains how it works differently: metal.equinix.com/proximity