GCP
Focus
Focus
AI Runtime Security

GCP

Table of Contents


GCP

AI Runtime Security post deployment configurations in Strata Cloud Manager (SCM) to protect VM workloads and Kubernetes clusters.
Where Can I Use This?What Do I Need?
  1. Log in to SCM.
  2. Configure AI Runtime Security instance (firewall) Interfaces:
    1. Select Manage → Configuration → NGFW and Prisma Access.
    2. Select Device Settings → Interfaces.
    3. Set the Configuration Scope to your AI Runtime Security folder.
    4. In Ethernet tab:
      Configure a Layer 3 Interface for eth1/1 and eth1/2:
      • Interfaces: eth1/1 and eth1/2
      • Location: Specify location if applicable
      • Interface Type: Layer3
      • IP Address: Dynamic (DHCP Client)
    5. Select the Loopback tab, to configure the Loopback interface:
      • In IPv4s, enter the ILB (Internal Load Balancer) private IP address
      • Set Security Zone to trust for eth1/2 and untrust for eth1/1
      • Ensure VR (Virtual Router) is set to default or the same as eth1/2
  3. Create zones. Select Manage → Configuration → NGFW and Prisma Access → Device Settings → Zones.
  4. Configure a Logical Router:
    • Create a Logical Router and add the Layer 3 interfaces (eth1/1 and eth1/2).
    • Configure a Static Route with the ILB static IP addresses for routing. Use the trust interface gateway IP address.
  5. Add a security policy (Manage → Configuration → NGFW and Prisma Access → Security Services → Security Policy → Add Rule).
    Ensure the policy allows health checks from the GCP Load Balancer (LB) pool to the internal LB IP from SCM. Check session IDs to ensure the firewall responds correctly on the designated interfaces.

Configurations to Secure VM Workloads

  1. Configure Static Routes for VPC endpoints.
    1. Log in to SCM.
    2. Select Manage → Configuration → NGFW and Prisma Access → Device Settings → Routing → Logical Routers.
    3. For VPC Subnet:
      • Edit the IPv4 Static Routes and add the route for the VPC IPv4 range CIDR subnets.
      • Set the Next Hop as eth1/2.
      • Set the Destination as the trust subnet gateway IP from SCM.
      • Update the static route.
      Save the Logical Router.
  2. Log in to SCM.
  3. Select Manage → Operations → Push Config and push the policy configurations to the AI Runtime Security instance.

Configurations to Secure the Kubernetes Clusters

  1. Add pod and service IP Subnets to AI Runtime Security trust firewall rules:
    1. Get the IP addresses for pod and service subnets:
    1. Go to Kubernetes Engine -> Clusters.
    2. Select a Cluster and copy the Cluster Pod IPv4 and IPv4 Service range IP addresses.
  2. Follow the AI Runtime Security Instance deployment in GCP to save and download the Terraform template.
  3. Edit the Terraform template to whitelist the following IP addresses in your VPC network firewall rules:
    • Navigate to the `<unzipped-folder>/architecture/security_project` directory.
    • Edit the `terraform.tfvars` file to add the copied IP addresses list to your `source_ranges`.
      firewall_rules = { allow-trust-ingress = { name = "allow-trust-vpc" source_ranges = ["35.xxx.0.0/16", "130.xxx.0.0/22", "192.xxx.0.0/16", "10.xxx.0.0/14", "10.xx.208.0/20"] # 1st 2 IPs are for health check packets. Add APP VPC/Pod/Service CIDRs priority = "1000" allowed_protocol = "all" allowed_ports = [] } }
  4. Apply the Terraform:
    terraform init terraform plan terraform apply
  5. Add static routes on the Logical Router for Kubernetes workloads:
    1. Log in to SCM.
    2. Select Manage → Configuration → NGFW and Prisma Access → Device Settings → Routing → Logical Routers.
    3. Configure Static Routes for the pod and service subnets for the Kubernetes workloads:
      Pod Subnet:
      • Edit the IPv4 Static Routes and add a route with the Pod IPv4 range CIDR.
      • Set the Next Hop as eth1/2 (trust interface).
      • Set the Destination as the trust subnet gateway IP from SCM.
      Service Subnet:
      • Edit the IPv4 Static Routes add a route with the IPv4 Service range CIDR.
      • Set the Next Hop as eth1/2 (trust interface).
      • Set the Destination as the trust subnet gateway IP from SCM.
  6. Add source NAT Policy for Outbound Traffic:
    1. Log in to SCM.
    2. Select Manage → Configuration → NGFW and Prisma Access → Network Policies → NAT.
    3. Create or modify a source NAT Policy:
      • Source Zone: Trust
      • Destination Zone: Untrust (eth1/1)
      • Policy Name: trust2untrust or similar.
  7. Configure NAT settings:
    Interface Address Section:
    • Set the Interface to eth1/1. (The translation happens at eth1/1).
      If needed, create a complementary rule for the reverse direction (for example, untrust2trust).
  8. Log in to SCM.
  9. Select Manage → Operations → Push Config and push the policy configurations to the AI Runtime Security instance.
    Note: If you have a Kubernetes cluster running, follow the section to install a Kubernetes application with Helm.

Install a Kubernetes Application with Helm

Follow the below steps to install a Kubernetes application on a Kubernetes cluster.
  1. Change the directory to the Helm folder:
    cd <unzipped-folder>/architecture/helm
  2. Create the `ai-runtime-security` directory and move the below files to this directory:
    mkdir ai-runtime-security mv Chart.yaml ai-runtime-security mv values.yaml ai-runtime-security mv templates ai-runtime-security
  3. Install the Helm chart:
    helm install ai-runtime-security ai-runtime-security --namespace kube-system --values ai-runtime-security/values.yaml
  4. Verify the Helm installation:
    #List all Helm releases helm list -A #Ensure the output shows your installation with details such as: NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION ai-runtime-security kube-system 1 2024-08-13 07:00 PDT deployed ai-runtime-security-0.1.0 11.2.2
  5. Check the pod status:
    kubectl get pods -A #Verify that the pods with names similar to `pan-cni-*****` are present.
  6. Check the endpoint slices:
    kubectl get endpointslice -n kube-system #Confirm that the output shows an ILB IP address: NAME ADDRESSTYPE PORTS ENDPOINTS AGE my-endpointslice IPv4 80/TCP 10.2xx.0.1,10.2xx.0.2 12h
  7. Check the services running in the `kube-system` namespace:
    kubectl get svc -n kube-system #Ensure that services `pan-cni-sa` and `pan-plugin-user-secret` are listed: NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE pan-cni-sa ClusterIP 10.xx.0.1 <none> 443/TCP 24h pan-plugin-user-secret ClusterIP 10.xx.0.2 <none> 443/TCP 24h
  8. Annotate the application `yaml` or `namespace` so that the traffic from the new pods is redirected to the AI Runtime Security instance (firewall) for inspection.
    annotations: paloaltonetworks.com/firewall: pan-fw
    For example, for all new pods in the "default" namespace:
    kubectl annotate namespace default paloaltonetworks.com/firewall=pan-fw