Splitting Service Traffic between K8S and VMs

In this how-to you’ll learn how to setup traffic routing between a service running both on a VM, and a Kubernetes cluster.

In this guide, you’ll
✓ Install the Istio demo bookinfo application in a cluster
✓ Install the ratings service from the bookinfo application on a VM.
✓ Split traffic 80/20 between the VM, and the cluster instances of the ratings application

Before you get started, make sure that you:
✓ Install TSB management plane
✓ Onboarded a cluster
✓ Install data plane operator

First, start by installing bookinfo in your cluster.

kubectl create ns bookinfo
kubectl apply -f \
    https://raw.githubusercontent.com/istio/istio/release-1.7/samples/bookinfo/platform/kube/bookinfo.yaml \
    -n bookinfo

Follow the VM onboarding docs. During onboarding, run the Istio demo ratings app as your workload.

sudo docker run -d \
    --name ratings \
    -p 127.0.0.1:9080:9080 \
    docker.io/istio/examples-bookinfo-ratings-v1.1:1.16.2

Create a workload entry for ratings,

apiVersion: networking.istio.io/v1beta1
kind: WorkloadEntry
metadata:
 name: ratings-vm
 namespace: bookinfo
 annotations:
   sidecar-bootstrap.istio.io/ssh-host: <ssh-host>
   sidecar-bootstrap.istio.io/ssh-user: istio-proxy
   sidecar-bootstrap.istio.io/proxy-config-dir: /etc/istio-proxy
   sidecar-bootstrap.istio.io/proxy-image-hub: docker.io/tetrate
   sidecar-bootstrap.istio.io/proxy-instance-ip: <proxy-instance-ip>
spec:
 address: <address>
 labels:
   class: vm
   app: ratings   # mandatory label for observability through TSB
   version: v3    # mandatory label for observability through TSB
 serviceAccount: bookinfo-ratings
 network: <vm-network-name>

and apply a sidecar.

apiVersion: networking.istio.io/v1beta1
kind: Sidecar
metadata:
  name: bookinfo-ratings-no-iptables
  namespace: bookinfo
spec:
  egress:
  - bind: 127.0.0.2
    hosts:
    - ./*
  ingress:
  - defaultEndpoint: 127.0.0.1:9080
    port:
      name: http
      number: 9080
      protocol: HTTP
  workloadSelector:
    labels:
      app: ratings
      class: vm

Once you've onboarded the VM, your mesh will distribute traffic between the ratings app in the cluster and the VM, because the ratings service selects any workload with the app: ratings label, and both our cluster Deployment and WorkloadEntry have this label. You can verify that traffic is flowing through both apps using logs or in the UI topology dashboard.

Now lets fine tune the traffic so that 80% of it goes to the in cluster app, and 20% goes to the VM. tctl apply -f a file containing the following configuration (filling in <tenant> and <cluster> based on your install).

note

You may already have a workspace set up (e.g. for ingress traffic). If that is the case, you can omit this workspace and adjust the rest of the config accordingly.

apiVersion: api.tsb.tetrate.io/v2
kind: Workspace
metadata:
  name: bookinfo-ws
  tenant: <tenant>
spec:
  namespaceSelector:
    names:
    - <cluster>/bookinfo
---
apiVersion: traffic.tsb.tetrate.io/v2
kind: Group
metadata:
  name: bookinfo-tg
  workspace: bookinfo-ws
  tenant: <tenant>
spec:
  namespaceSelector:
    names:
      - "<cluster>/bookinfo"
  configMode: BRIDGED
---
apiVersion: traffic.tsb.tetrate.io/v2
kind: ServiceRoute
metadata:
  name: ratings
  group: bookinfo-tg
  workspace: bookinfo-ws
  tenant: <tenant>
spec:
  service: bookinfo/ratings
  subsets:
  - name: v1
    labels:
      version: v1
    weight: 80
  - name: v3
    labels:
      version: v3
    weight: 20

After sending some traffic through the app, we can look at the services dashboard or logs again to see that the traffic is being split 80/20 between ratings v1 and v3.