faas-netes - Serverless Functions For Kubernetes with OpenFaaS

Introduction

faas-netes is an OpenFaaS provider which enables Kubernetes for OpenFaaS. It's part of a larger stack that brings a cloud-agnostic serverless experience to Kubernetes.

The existing REST API, CLI and UI are fully compatible. With OpenFaaS Pro, you have an optional operator mode so that you can manage functions with kubectl and a CustomResource.

You can deploy OpenFaaS to any Kubernetes service - whether managed or local, including to OpenShift. You will find any specific instructions and additional links in the documentation.

OpenFaaS (Functions as a Service) is a framework for building serverless functions with Docker and Kubernetes which has first class support for metrics. Any process can be packaged as a function enabling you to consume a range of web events without repetitive boiler-plate coding.

Pictured: OpenFaaS conceptual architecture

Highlights

Platform for deploying serverless-style workloads - microservices and functions
Native Kubernetes integrations (API and ecosystem)
Built-in UI portal
Scale to and from zero
Built-in queuing and asynchronous invocations
Custom routes and domain support
Commercial support available

Additional & ecosystem:

A range of event-connectors and cron-support
helm chart and CLI installer
Operator available to use Custom Resource Definitions (CRDs) (See also: OpenFaaS Pro)
IDp integration with OIDC and commercial add-on

Community:

Over 30k GitHub stars
Independent open-source project with over 300 contributors, with commercial option available

Commercial options:

Support from full-time team
Commercial add-ons and integrations with events like Kafka, Postgres, AWS SQS and Cron
Multiple namespace support
gVisor support and runtimeClass for isolation
Affinity and advanced scheduling / security constraints

Get started

The PLONK Stack

OpenFaaS can be used as complete stack for Cloud Native application development called PLONK. The PLONK Stack includes: Prometheus, Linux/Linkerd, OpenFaaS, NATS/Nginx and Kubernetes.

Technical and operational information

The rest of this document is dedicated to technical and operational information for the controller.

Operating modes - classic or operator

There are two modes available for faas-netes, the classic mode is the default.

Classic mode (aka faas-netes) - includes a REST API, multiple-namespace support but no Function CRD - available in Community Edition and OpenFaaS Pro/Enteprise
Operator mode (aka "The OpenFaaS Operator") - includes a REST API, with a "Function" CRD and multiple-namespace OpenFaaS Pro/Enterprise

The single faas-netes image and binary contains both modes, switch between one or the other using the helm chart or the flag -operator=true/false.

Configuration of the controller

faas-netes can be configured with environment variables, but for a full set of options see the helm chart.

Option	Usage
`httpProbe`	Boolean - use http probe type for function readiness and liveness. Default: `true`
`write_timeout`	HTTP timeout for writing a response body from your function (in seconds). Default: `60s`
`read_timeout`	HTTP timeout for reading the payload from the client caller (in seconds). Default: `60s`
`image_pull_policy`	Image pull policy for deployed functions (`Always`, `IfNotPresent`, `Never`). Default: `Always`
`gateway.resources`	CPU/Memory resources requests/limits (memory: `120Mi`, cpu: `50m`)
`faasnetes.resources`	CPU/Memory resources requests/limits (memory: `120Mi`, cpu: `50m`)
`operator.resources`	CPU/Memory resources requests/limits (memory: `120Mi`, cpu: `50m`)
`queueWorker.resources`	CPU/Memory resources requests/limits (memory: `120Mi`, cpu: `50m`)
`prometheus.resources`	CPU/Memory resources requests/limits (memory: `512Mi`)
`alertmanager.resources`	CPU/Memory resources requests/limits (memory: `25Mi`)
`nats.resources`	CPU/Memory resources requests/limits (memory: `120Mi`)
`faasIdler.resources`	CPU/Memory resources requests/limits (memory: `64Mi`)
`basicAuthPlugin.resources`	CPU/Memory resources requests/limits (memory: `50Mi`, cpu: `20m`)

Readiness checking

The readiness checking for functions assumes you are using our function watchdog which writes a .lock file in the default "tempdir" within a container. To see this in action you can delete the .lock file in a running Pod with kubectl exec and the function will be re-scheduled.

Namespaces

By default all OpenFaaS functions and services are deployed to the openfaas and openfaas-fn namespaces. To alter the namespace use the helm chart.

Ingress

To configure ingress see the helm chart. By default NodePorts are used. These are listed in the deployment guide.

By default functions are exposed at http://gateway:8080/function/NAME.

You can also use the IngressOperator to set up custom domains and HTTP paths

Image pull policy

By default, deployed functions will use an imagePullPolicy of Always, which ensures functions using static image tags are refreshed during an update.

If this is not desired behavior, OpenFaaS Pro customers can set the image_pull_policy environment variable to an alternative. IfNotPresent is particularly useful when developing locally with minikube. In this case, you can set your local environment to use minikube's docker so faas-cli build builds directly into minikube's image store. faas-cli push is unnecessary in this workflow - use faas-cli build then faas-cli deploy.

Note: When set to Never, only local (or pulled) images will work. When set to IfNotPresent, function deployments may not be updated when using static image tags.

Kubernetes Versions

faas-netes maintainers strive to support as many Kubernetes versions as possible and it is currently compatible with Kubernetes 1.11 and higher. Instructions for OpenShift are also available in the documentation.