Network Monitoring

Network access traces with Tetragon

In addition to file access monitoring, Tetragon’s tracing policies also support monitoring network access. In this section, you will see how to monitor network traffic to “external” destinations (destinations that are outside the Kubernetes cluster or external to the Docker host where Tetragon is running). These instructions assume you already have Tetragon running in either Kubernetes or Docker, and that you have deployed the Cilium demo application.

Monitoring Kubernetes network access

First, you’ll need to find the pod CIDR and service CIDR in use. In many cases the pod CIDR is relatively easy to find.

export PODCIDR=`kubectl get nodes -o jsonpath='{.items[*].spec.podCIDR}'`

You can fetch the service CIDR from the cluster in some environments. When working with managed Kubernetes offerings (AKS, EKS, or GKE) you will need the environment variables used when you created the cluster.

export SERVICECIDR=$(gcloud container clusters describe ${NAME} --zone ${ZONE} | awk '/servicesIpv4CidrBlock/ { print $2; }')
export SERVICECIDR=$(kubectl describe pod -n kube-system kube-apiserver-kind-control-plane | awk -F= '/--service-cluster-ip-range/ {print $2; }')
export SERVICECIDR=$(aws eks describe-cluster --name ${NAME} | jq -r '.cluster.kubernetesNetworkConfig.serviceIpv4Cidr')
export SERVICECIDR=$(az aks show --name ${NAME} --resource-group ${AZURE_RESOURCE_GROUP} | jq -r '.networkProfile.serviceCidr)

Once you have this information, you can customize a policy to exclude network traffic to the networks stored in the PODCIDR and SERVICECIDR environment variables. Use envsubst to do this, and then apply the policy to your Kubernetes cluster with kubectl apply:

wget https://raw.githubusercontent.com/cilium/tetragon/main/examples/quickstart/network_egress_cluster.yaml
envsubst < network_egress_cluster.yaml | kubectl apply -f -

Once the tracing policy is applied, you can attach tetra to observe events again:

kubectl exec -ti -n kube-system ds/tetragon -c tetragon -- tetra getevents -o compact --pods xwing --processes curl
POD=$(kubectl -n kubesystem get pods -l 'app.kubernetes.io/name=tetragon' -o name --field-selector spec.nodeName=$(kubectl get pod xwing -o jsonpath='{.spec.nodeName}'))
kubectl exec -ti -n kube-system $POD -c tetragon -- tetra getevents -o compact --pods xwing --processes curl

Then execute a curl command in the “xwing” Pod to access one of our favorite sites.

 kubectl exec -ti xwing -- bash -c 'curl https://ebpf.io/applications/#tetragon'

You will observe a connect event being reported in the output of the tetra getevents command:

🚀 process default/xwing /usr/bin/curl https://ebpf.io/applications/#tetragon
🔌 connect default/xwing /usr/bin/curl tcp 10.32.0.19:33978 -> 104.198.14.52:443
💥 exit    default/xwing /usr/bin/curl https://ebpf.io/applications/#tetragon 60

You can confirm in-kernel BPF filters are not producing events for in-cluster traffic by issuing a curl to one of our services and noting there is no connect event.

kubectl exec -ti xwing -- bash -c 'curl -s -XPOST deathstar.default.svc.cluster.local/v1/request-landing'

The output should be similar to:

Ship landed

And as expected no new events:

🚀 process default/xwing /usr/bin/curl https://ebpf.io/applications/#tetragon
🔌 connect default/xwing /usr/bin/curl tcp 10.32.0.19:33978 -> 104.198.14.52:443
💥 exit    default/xwing /usr/bin/curl https://ebpf.io/applications/#tetragon 60

Monitoring Docker or bare metal network access

This example also works easily for local Docker users. However, since Docker does not have pod CIDR or service CIDR constructs, you will construct a tracing policy that filters 127.0.0.1 from the Tetragon event log.

First, set the necessary environment variables to the loopback IP address.

export PODCIDR="127.0.0.1/32"
export SERVICECIDR="127.0.0.1/32"

Next, customize the policy using envsubst.

wget https://raw.githubusercontent.com/cilium/tetragon/main/examples/quickstart/network_egress_cluster.yaml
envsubst < network_egress_cluster.yaml > network_egress_cluster_subst.yaml

Finally, start Tetragon with the new policy.

docker stop tetragon
docker run -d --name tetragon --rm --pull always \
  --pid=host --cgroupns=host --privileged               \
  -v ${PWD}/network_egress_cluster_subst.yaml:/etc/tetragon/tetragon.tp.d/network_egress_cluster_subst.yaml \
  -v /sys/kernel/btf/vmlinux:/var/lib/tetragon/btf      \
  quay.io/cilium/tetragon:v1.2.0

Once Tetragon is running, use docker exec to run the tetra getevents command and log the output to your terminal.

docker exec -ti tetragon tetra getevents -o compact

Now remote TCP connections will be logged, but connections to the localhost address are filtered out by Tetragon. You can see this by executing a curl command to generate a remote TCP connect.

curl https://ebpf.io/applications/#tetragon

This produces the following output:

🚀 process  /usr/bin/curl https://ebpf.io/applications/#tetragon
🔌 connect  /usr/bin/curl tcp 192.168.1.190:36124 -> 104.198.14.52:443
💥 exit     /usr/bin/curl https://ebpf.io/applications/#tetragon 0

What’s next

So far you have installed Tetragon and used a couple policies to monitor sensitive files and provide network auditing for connections outside your own cluster and node. Both these cases highlight the value of in-kernel filtering. Another benefit of in-kernel filtering is you can add enforcement to the policies to not only alert via a log entry, but to block the operation in kernel and/or kill the application attempting the operation.

To learn more about policies and events Tetragon can implement review the Concepts section.