Skip to content

Deploying Wallarm Sidecar Proxy

To secure an application deployed as a Pod in a Kubernetes cluster, you can run the NGINX-based Wallarm node in front of the application as a sidecar controller. Wallarm sidecar controller will filter incoming traffic to the application Pod by allowing only legitimate requests and mitigating malicious ones.

The key features of the Wallarm Sidecar proxy solution:

  • Simplifies protection of discrete microservices and their replicas and shards by providing the deployment format that is similar to applications

  • Fully compatible with any Ingress controller

  • Works stable under high loads that is usually common for the service mesh approach

  • Requires minimum service configuration to secure your apps; just add some annotations and labels for the application pod to protect it

  • Supports two modes of the Wallarm container deployment: for medium loads with the Wallarm services running in one container and for high loads with the Wallarm services split into several containers

  • Provides a dedicated entity for the postanalytics module that is the local data analytics backend for the Wallarm sidecar proxy solution consuming most of the CPU

If you are using the earlier Wallarm Sidecar solution

If you are using the previous version of the Wallarm Sidecar solution, we recommend you migrate to the new one. With this release, we updated our Sidecar solution to leverage new Kubernetes capabilities and a wealth of customer feedback. The new solution does not require significant Kubernetes manifest changes, to protect an application, just deploy the chart and add labels and annotations to the pod.

For assistance in migrating to the Wallarm Sidecar proxy solution v2.0, please contact Wallarm technical support.

Use cases

Among all supported Wallarm deployment options, this solution is the recommended one for the following use cases:

  • You are looking for the security solution to be deployed to the infrastructure with the existing Ingress controller (e.g. AWS ALB Ingress Controller) preventing you from deployment of either Wallarm NGINX-based or Wallarm Kong-based Ingress controller

  • Zero-trust environment that requires each microservice (including internal APIs) to be protected by the security solution

  • The security solution should allow pods to reach VPCs to access your APIs

  • The security solution should be compatible with third-party services routing your traffic like AWS API Gateway

Traffic flow

Traffic flow with Wallarm Sidecar proxy:

Traffic flow with Wallarm Sidecar proxy

Solution architecture

The Wallarm Sidecar proxy solution is arranged by the following Deployment objects:

  • Sidecar controller (wallarm-sidecar-controller) is the mutating admission webhook that injects Wallarm sidecar proxy resources into the Pod configuring it based on the Helm chart values and pod annotations and connecting the node components to the Wallarm Cloud.

    Once a new pod with the wallarm-sidecar: enabled label in Kubernetes starts, the controller automatically injects the additional container filtering incoming traffic into the pod.

  • Postanalytics module (wallarm-sidecar-postanalytics) is the local data analytics backend for the Wallarm sidecar proxy solution. The module uses the in-memory storage Tarantool and the set of some helper containers (like the collectd, attack export services).

Wallarm deployment objects

The Wallarm Sidecar proxy has 2 standard stages in its lifecycle:

  1. At the initial stage, the controller injects Wallarm sidecar proxy resources into the Pod configuring it based on the Helm chart values and pod annotations and connecting the node components to the Wallarm Cloud.

  2. At the runtime stage, the solution analyzes and proxies/forwards requests involving the postanalytics module.

Requirements

  • Kubernetes platform version 1.19-1.25

  • Helm v3 package manager

  • An application deployed as a Pod in a Kubernetes cluster

  • Access to https://us1.api.wallarm.com for working with US Wallarm Cloud or to https://api.wallarm.com for working with EU Wallarm Cloud

  • Access to https://charts.wallarm.com to add the Wallarm Helm charts

  • Access to the Wallarm repositories on Docker Hub https://hub.docker.com/r/wallarm

  • Access to the IP addresses of Google Cloud Storage listed within the link. When you allowlist, denylist, or graylist entire countries, regions, or data centers instead of individual IP addresses, the Wallarm node retrieves precise IP addresses related to the entries in the IP lists from the aggregated database hosted on Google Storage

  • Access to the account with the Administrator role in Wallarm Console for the US Cloud or the EU Cloud

Deployment

To deploy the Wallarm Sidecar proxy solution:

  1. Create the Wallarm node.

  2. Deploy the Wallarm Helm chart.

  3. Attach the Wallarm Sidecar proxy to the application Pod.

  4. Test the Wallarm Sidecar proxy operation.

Step 1: Create the Wallarm node

  1. Open Wallarm Console → Nodes via the link below:

  2. Create a filtering node with the Wallarm node type and copy the generated token.

    Creation of a Wallarm node

Step 2: Deploy the Wallarm Helm chart

  1. Add the Wallarm chart repository:

    helm repo add wallarm https://charts.wallarm.com
    

  2. Create the values.yaml file with the Wallarm Sidecar proxy configuration.

    Example of the file with the minimum configuration:

    config:
      wallarm:
        api:
          token: "<NODE_TOKEN>"
          host: "us1.api.wallarm.com"
    
    config:
      wallarm:
        api:
          token: "<NODE_TOKEN>"
    

    <NODE_TOKEN> is the token of the Wallarm node to be run in Kubernetes.

    Using one token for several installations

    You can use one token in several installations regardless of the selected platform. It allows logical grouping of node instances in the Wallarm Console UI. Example: you deploy several Wallarm nodes to a development environment, each node is on its own machine owned by a certain developer.

  3. Deploy the Wallarm Helm chart:

    helm install --version 4.6.5 <RELEASE_NAME> wallarm/wallarm-sidecar --wait -n wallarm-sidecar --create-namespace -f <PATH_TO_VALUES>
    
    • <RELEASE_NAME> is the name for the Helm release of the Wallarm Sidecar proxy chart
    • wallarm-sidecar is the new namespace to deploy the Helm release with the Wallarm Sidecar proxy chart, it is recommended to deploy it to a separate namespace
    • <PATH_TO_VALUES> is the path to the values.yaml file

Step 3: Attach the Wallarm Sidecar proxy to the application Pod

For Wallarm to filter application traffic, add the wallarm-sidecar: enabled label to the corresponding application Pod:

kubectl edit deployment -n <KUBERNETES_NAMESPACE> <APP_LABEL_VALUE>
apiVersion: apps/v1
kind: Deployment
metadata:
  name: myapp
  namespace: default
spec:
  replicas: 1
  selector:
    matchLabels:
      app: myapp
  template:
    metadata:
      labels:
        app: myapp
        wallarm-sidecar: enabled
    spec:
      containers:
        - name: application
          image: kennethreitz/httpbin
          ports:
            - name: http
              containerPort: 80
  • If the wallarm-sidecar application Pod label is either set to disabled or not explicitly specified, the Wallarm Sidecar container is not injected into a pod and therefore Wallarm does not filter traffic.

  • If the wallarm-sidecar application Pod label is set to enabled, the Wallarm Sidecar container is injected into a pod and therefore Wallarm filters incoming traffic.

Step 4: Test the Wallarm Sidecar proxy operation

To test that the Wallarm Sidecar proxy operates correctly:

  1. Get the Wallarm pod details to check they have been successfully started:

    kubectl get pods -n <NAMESPACE> -l app.kubernetes.io/name=wallarm-sidecar
    

    Each pod should display the following: READY: N/N and STATUS: Running, e.g.:

    NAME                                              READY   STATUS    RESTARTS   AGE
    wallarm-sidecar-controller-54cf88b989-f7jtb      1/1     Running   0          91m
    wallarm-sidecar-controller-54cf88b989-gp2vg      1/1     Running   0          91m
    wallarm-sidecar-postanalytics-86d9d4b6cd-hpd5k   4/4     Running   0          91m
    
  2. Get the application pod details to check the Wallarm sidecar controller has been successfully injected:

    kubectl get pods --selector app=<APP_LABEL_VALUE>
    

    The output should display READY: 2/2 pointing to successful sidecar container injection and STATUS: Running pointing to successful connection to the Wallarm Cloud:

    NAME                     READY   STATUS    RESTARTS   AGE
    myapp-5c48c97b66-lzkwf   2/2     Running   0          3h4m
    
  3. Send the test Path Traversal attack to the application cluster address Wallarm is enabled to filter traffic:

    curl http://<APPLICATION_CLUSTER_IP>/etc/passwd
    

    Since the Wallarm proxy operates in the monitoring filtration mode by default, the Wallarm node will not block the attack but will register it.

    To check that the attack has been registered, proceed to Wallarm Console → Events:

    Attacks in the interface

Customization

Wallarm pods have been injected based on the default values.yaml and the custom configuration you specified on the 2nd deployment step.

You can customize the Wallarm proxy behavior even more on both the global and per-pod levels and get the most out of the Wallarm solution for your company.

Just proceed to the Wallarm proxy solution customization guide.