1,981 reads

How To Deploy Firecracker MicroVMs using Weave FireKube

by Sudip SenguptaAugust 23rd, 2020

Too Long; Didn't Read

How To Deploy Firecracker MicroVMs using Weave FireKube is a Kubernetes cluster working on top of Weave Ignite and FireCracker. Using the same technology that Amazon uses for AWS Lambda and AWS Fargate, Amazon announced an exciting new product called "Firecracker" that is quickly setting the cloud-native ecosystem on fire. Using Ignite, you just pick a container image that you want to run as a VM and then execute it. With GitOps, the cluster is then running inside containers.

Companies Mentioned

featured image - How To Deploy Firecracker MicroVMs using Weave FireKube

Introduction

At their 2018 annual Re:Invent conference, AWS announced an exciting new product called "Firecracker" that is quickly setting the cloud-native ecosystem on fire. Firecracker is a Virtual Machine Manager (VMM) exclusively designed for running transient and short-lived processes. In other words, it is optimized for running functions and serverless workloads that require faster cold start and higher density.

Using the same technology that Amazon uses for AWS Lambda and AWS Fargate, Firecracker delivers the speed of containers with the security of VMs and has the potential to disrupt the current container and serverless technologies.

What is Weave Ignite?

While Firecracker has a ton of potential, it's still in its early stages so getting it up and running can be a bit tedious. The goal of Ignite is to solve this issue by adopting its developer experience from containers.

With Ignite, you just pick a container image that you want to run as a VM and then execute

ignite run

instead of

docker run

.

There’s no need to use VM-specific tools to build

.vdi

.vmdk

, or

.qcow2

images, just do a

docker build

from any base image you want (e.g.

ubuntu:18.04

from Docker Hub), and add your preferred content.

You can even use Buildpacks! As seen in my other blog post here. In this case, you would just follow the same steps, except you would build the image via the

pack

command.

What is Firekube?

Firekube is a new open-source Kubernetes distribution that enables the use of Weave Ignite and GitOps to enable the setup of secure VM clusters. Firekube pulls everything from Git, detects your operating system and can boot up a secure cluster of VMs from nothing in 2.5 minutes.

A Firekube cluster has the following properties:

Runs Kubernetes (now K8s, possibly K3s in future)
High-grade VM security via the Firecracker KVM isolation
Fast start-up and tear down of VMs e.g. for functions and serverless apps
Scales from zero to production - uses standard k8s plugins for networking, etc
“Lift and shift” software into VMs
Run containers inside VMs or alongside VMs on the same CNI network

Creating a Firekube Cluster

Firekube is a Kubernetes cluster working on top of Ignite and Firecracker. Firekube clusters are operated with GitOps.

Ignite and Firecracker only works on Linux as they need KVM. However, it will also work on macOS using footloose: the Kubernetes nodes are then running inside containers.

Prerequisites: Docker, Git, kubectl 1.14+.

1. Fork this repository.

2. Clone your fork and

cd

into it. Use the

SSH

git URL as the script will push an initial commit to your fork:

export user="" # Your GitHub handle or org

git clone [email protected]:$user/wks-quickstart-firekube.git
cd wks-quickstart-firekube

3. Start the cluster:

./setup.sh

This step will take several minutes.

Export the KUBECONFIG environment variable as indicated at the end of the installation:

export KUBECONFIG=/home/damien/.wks/weavek8sops/example/kubeconfig

Enjoy your Kubernetes cluster!

$ kubectl get nodes
NAME               STATUS   ROLES    AGE     VERSION
67bb6c4812b19ce4   Ready    master   3m42s   v1.14.1
a5cf619fa058882d   Ready    <none>   75s     v1.14.1

Watch GitOps in action

Now that we have a cluster installed, we can commit Kubernetes objects to the git repository and have them appear in the cluster. Let's add podinfo, an example Go microservice, to the cluster.

kubectl apply --dry-run -k github.com/stefanprodan/podinfo//kustomize -o yaml > podinfo.yaml
git add podinfo.yaml
git commit -a -m 'Add podinfo Deployment'
git push

A few seconds later, you should witness the apparition of a podinfo pod in the cluster:

$ kubectl get pods
NAME                       READY   STATUS    RESTARTS   AGE
podinfo-677768c755-z76xk   1/1     Running   0          30s

To view podinfo web UI:

1. Expose

podinfo

locally:

kubectl port-forward deploy/podinfo 9898:9898

2. Point your browser to

http://127.0.0.1:9898

Deleting a Firekube cluster

Run:

./cleanup.sh

Using a private git repository with firekube

To use a private git repository instead of a fork of

wks-quickstart-firekube

1. Create a private repository and push the

wks-quickstart-firekube

master

branch there. Use the SSH git URL when cloning the private repository:

git clone [email protected]:$user/$repository.git
cd $repository
git remote add quickstart [email protected]:weaveworks/wks-quickstart-firekube.git
git fetch quickstart
git merge quickstart/master
git push

2. Create an SSH key pair:

ssh-keygen -t rsa -b 4096 -C "[email protected]" -f deploy-firekube  -N ""

3. Upload the deploy key to your private repository (with read/write access):

4. Start the cluster:

./setup.sh --git-deploy-key  ./deploy-firekube

Final thoughts

You've now learned the basic steps it requires to provision a Kubernetes cluster based on Firecracker VM's with Firekube. You should explore a bit of the different options the platform provides such as using Ignite to spin up MicroVM's with Buildpack instead of Docker. I hope you liked this post, and I plan to dive into this a bit more in the future. Thanks for reading!

About the author - Sudip is a Solution Architect with more than 15 years of working experience, and is the founder of Javelynn. He likes sharing his knowledge by regularly writing for Hackernoon, DZone, Appfleet and many more. And while he is not doing that, he must be fishing or playing chess.

Previously posted at https://appfleet.com/.