antojoseph/fc

Fuzzing Infrastructure with k8s & cephfs

Distributed fuzzing infrastructure with k8s, ceph for state sharing and elk stack for monitoring!

• scale your fuzzing workloads
• resilient sate sharing using ceph
• cluster performace monitoring with elastic,metric-beats and kibana

To-Do

• Squash Container Image

Planned Features!

• Support afl++
• Support fuzzilli (docker)
• CI/CD system for continous fuzzing

Tech

FuzzCube uses a number of open source projects:

• afl - the mighty afl
• libFuzzer - powerful in-process fuzzing
• k8s - Container Orchestration!
• ceph - Distributed Object Storage
• elastic - Visibility for k8s.

And of course FuzzCube itself is open source with a public repository on GitHub.

Test Cluster using minikube

FuzzCube requires a working k8s cluster to run. Install minikube and deploy the yaml spec to start fuzzing.

#install minikube
$brew install minikube
#or to update your existing installation
$brew update
$brew upgrade minikube

#deploy a k8s cluster with 4 cpu , 8gigs of memory , 20 gigs of storage and log startup to console
$minikube -p clustername start --cpus=4 --memory=8192 --disk-size 20GB --alsologtostderr -v=7

# deploy dashboard for a neat UI to see everything in your cluster
$minikube dashboard -p clustername

#deploy metrics-server required by k8s to measure performance
$minikube addons -p clustername metrics-server

# to view resource utilization in the cluster
$kubectl top pod --all-namespaces | sort --reverse --key 3 --numeric 

#to check what cluster you are working with 
$kubectl config get-contexts

# get all pods in the cluster ( all -namespaces )
$kubectl get pods -A

# get a shell inside a pod inside the cluster (for debugging ) 
$kubectl exec -it podnamehere /bin/bash

Installation

to deploy the fuzzer , have kubectl configured to use the cluster of choice . If you follow the above steps , this will be done for you!

$ git clone https://github.com/antojoseph/fc
$ cd fc
$ kubectl apply -f spellbook.yaml
$ kubectl apply -f afl.yml
$ kubectl apply -f libfuzzer.yml
$ kubectl apply -f debug.yml

For production environments...

$ change your ceph configuration to multi-node for resilience

Debugging

Use hyperkit driver for minikube if you are running OSX

$ minikube start --driver=hyperkit --cpus=4 --memory=8192 --disk-size 20GB --alsologtostderr -v=7

Youtube Demo

Credits

• saadrana219 - for sharing your wisdom

License

Apache License 2.0

Free Software, Hell Yeah!

fc's Issues

Missing Docker Image

Hello, I'm a student trying to run ClusterFuzz in a parallel way. I recently run your project, but it seems missing Docker Image: antojoseph/libfuzzer.
How can I get it or I should create my owns? Do you have any suggestion?
Thanks for your time.