Chore: Bring Linkerd GAMMA HTTPRoute example up to date

Author: baburciu

examples/linkerd-header-based/README.md

@@ -6,28 +6,29 @@
 
 A Kubernetes cluster. If you do not have one, you can create one using [kind](https://kind.sigs.k8s.io/), [minikube](https://minikube.sigs.k8s.io/), or any other Kubernetes cluster. This guide will use Kind.
 
-Linkerd installed in your Kubernetes cluster.
-
-
-## Step 1 - Create a Kind cluster by running the following command:
+## Step 1 - Create a Kind cluster by running the following command
 
 ```shell
-kind delete cluster &>/dev/null
 kind create cluster --config ./kind-cluster.yaml
 ```
 
-## Step 2 - Install Linkerd and Linkerd Viz by running the following commands:
+## Step 2 - Install Linkerd and Linkerd Viz by running the following commands
 
 I will use the Linkerd CLI to install Linkerd in the cluster. You can also install Linkerd using Helm or kubectl.
-I tested this guide with Linkerd version 2.14.10
+I tested this guide with Linkerd version `edge-25.9.4`.
+
+> [!IMPORTANT]
+> Linkerd version `edge-25.9.4` uses `v1` GatewayAPI apiVersion and the [plugin](https://github.com/argoproj-labs/rollouts-plugin-trafficrouter-gatewayapi) `v0.8.0` expects that. It wouldn't work if `v1beta1` GatewayAPI apiVersion CRDs would be installed (like in the case of an older Linkerd `stable-2.14.10`)
 
 ```shell
+export LINKERD2_VERSION=edge-25.9.4; curl --proto '=https' --tlsv1.2 -sSfL https://run.linkerd.io/install-edge | sh
+export PATH=$PATH:/home/baburciu/.linkerd2/bin
+kubectl apply -f https://github.com/kubernetes-sigs/gateway-api/releases/download/v1.2.1/standard-install.yaml # Gateway API CRDs must be installed prior to installing Linkerd
 linkerd install --crds | kubectl apply -f -
 linkerd install | kubectl apply -f - && linkerd check
 linkerd viz install | kubectl apply -f - && linkerd check
 ```
 
-
 ## Step 3 - Install Argo Rollouts and Argo Rollouts plugin to allow Linkerd to manage the traffic:
 
 ```shell
@@ -38,7 +39,7 @@ kubectl apply -f argo-rollouts-plugin.yaml
 kubectl rollout restart deploy -n argo-rollouts
 ```
 
-## Step 4 - Grant Argo Rollouts SA access to the Gateway/Http Route
+## Step 4 - Grant Argo Rollouts SA access to the HTTPRoute
 ```shell
 kubectl apply -f cluster-role.yaml
 ```
@@ -51,22 +52,23 @@ kubectl apply -f cluster-role-binding.yaml
 ```
 ## Step 5 - Create HTTPRoute that defines a traffic split between two services
 
-Create HTTPRoute and connect to the created Gateway resource
+Create a GAMMA [producer `HTTPRoute`](https://gateway-api.sigs.k8s.io/concepts/glossary/#producer-route) resource and connect it to a parent K8s service (using a canary and stable K8s services as backends)
 
 ```shell
 kubectl apply -f httproute.yaml
 ```
-## Step 6 - Create the services required for traffic split 
 
-Create three Services required for canary based rollout stratedy
+## Step 6 - Create the services required for traffic split
+
+Create three Services required for canary based rollout strategy
 
 ```shell
 kubectl apply -f service.yaml
 ```
 
-## Step 7 - Create the services required for traffic split 
+## Step 7 - Add `linkerd.io/inject: enabled` annotation to namespace
 
-Add Linkerd annotation to the namespace where the services are deployed
+Add Linkerd annotation to the namespace where the pods are deployed to enable [Automatic Proxy Injection](https://linkerd.io/2-edge/features/proxy-injection/)
 
 ```shell
 kubectl apply -f namespace.yaml
@@ -75,12 +77,15 @@ kubectl apply -f namespace.yaml
 ## Step 8 - Create an example Rollout
 
 Deploy a rollout to get the initial version.
+
 ```shell
 kubectl apply -f rollout.yaml
 ```
 
 ## Step 9 - Watch the rollout
-Monitor the HTTPRoute configuration to see how traffic is split and header-based routing is configured:
+
+Monitor the HTTPRoute configuration to see how traffic is split and header-based routing is configured
+
 ```shell
 watch "kubectl get httproute.gateway.networking.k8s.io/argo-rollouts-http-route -o jsonpath='{\" HEADERS: \"}{.spec.rules[*].matches[*]}'"
 ```
@@ -90,16 +95,102 @@ watch "kubectl get httproute.gateway.networking.k8s.io/argo-rollouts-http-route
 kubectl patch rollout rollouts-demo --type='json' -p='[{"op": "replace", "path": "/spec/template/spec/containers/0/env/0/value", "value": "1.1.0"}]'
 ```
 
-## Step 11 - Test the header-based routing with curl
-
-You can test the header-based routing by sending requests with and without the specified header:
+## Step 11 - Observe the rollout and HTTPRoute rule addition of [canary header matching rule](https://gateway-api.sigs.k8s.io/guides/traffic-splitting/#canary-traffic-rollout)
 
-# Without header (goes to stable)
 ```shell
-curl http://localhost:80
+$ kubectl argo rollouts promote rollouts-demo  # promote to Rollout step 1
+$ kubectl argo rollouts get rollout rollouts-demo
+Name:            rollouts-demo
+Namespace:       default
+Status:          ॥ Paused
+Message:         CanaryPauseStep
+Strategy:        Canary
+  Step:          3/5
+  SetWeight:     0
+  ActualWeight:  0
+Images:          hashicorp/http-echo (canary, stable)
+Replicas:
+  Desired:       5
+  Current:       6
+  Updated:       1
+  Ready:         6
+  Available:     6
+
+NAME                                       KIND        STATUS     AGE    INFO
+⟳ rollouts-demo                            Rollout     ॥ Paused   4m53s
+├──# revision:2
+│  └──⧉ rollouts-demo-8598f766fd           ReplicaSet  ✔ Healthy  67s    canary
+│     └──□ rollouts-demo-8598f766fd-2cvt4  Pod         ✔ Running  31s    ready:2/2
+└──# revision:1
+   └──⧉ rollouts-demo-5d78c448f9           ReplicaSet  ✔ Healthy  4m53s  stable
+      ├──□ rollouts-demo-5d78c448f9-9zttx  Pod         ✔ Running  4m53s  ready:2/2
+      ├──□ rollouts-demo-5d78c448f9-hpcfg  Pod         ✔ Running  4m53s  ready:2/2
+      ├──□ rollouts-demo-5d78c448f9-l757w  Pod         ✔ Running  4m53s  ready:2/2
+      ├──□ rollouts-demo-5d78c448f9-xl72c  Pod         ✔ Running  4m53s  ready:2/2
+      └──□ rollouts-demo-5d78c448f9-zk5pd  Pod         ✔ Running  4m53s  ready:2/2
+$
+$ kubectl get httproute argo-rollouts-http-route -o yaml | yq .spec.rules
+- backendRefs:
+    - group: core
+      kind: Service
+      name: argo-rollouts-stable-service
+      port: 80
+      weight: 100
+    - group: core
+      kind: Service
+      name: argo-rollouts-canary-service
+      port: 80
+      weight: 0
+  matches:
+    - path:
+        type: PathPrefix
+        value: /
+- backendRefs:
+    - group: ""
+      kind: Service
+      name: argo-rollouts-canary-service
+      port: 80
+      weight: 0
+  matches:
+    - headers:
+        - name: X-Test
+          type: Exact
+          value: test
+      path:
+        type: PathPrefix
+        value: /
+```
+```shell
+$ kubectl run -it --image nicolaka/netshoot:v0.13 network-test -- sh  # run a pod to source curl tests
+$ kubectl exec -it network-test -c network-test -- sh
+~ # # stable K8s service targets any of the 5 stable pods
+~ # curl http://argo-rollouts-stable-service/
+Hello from rollouts-demo-5d78c448f9-l757w
+~ #
+~ # # canary K8s service targets the one canary pod, created for the `setCanaryScale` step in the Rollout
+~ # curl http://argo-rollouts-canary-service/
+Hello from rollouts-demo-8598f766fd-2cvt4
+~ #
+~ # # GAMMA-type HTTPRoute's `.spec.parentRefs` K8s service only targets stable pods since no `setWeight` step is used in the Rollout
+~ # seq 1 100 | xargs -P 10 -I {} bash -c 'curl -s http://argo-rollouts-service' > pods.txt
+~ # sort pods.txt | uniq -c | sort -rn
+     25 Hello from rollouts-demo-5d78c448f9-hpcfg
+     21 Hello from rollouts-demo-5d78c448f9-xl72c
+     19 Hello from rollouts-demo-5d78c448f9-zk5pd
+     19 Hello from rollouts-demo-5d78c448f9-l757w
+     16 Hello from rollouts-demo-5d78c448f9-9zttx
+~ #
 ```
 
-# With header (goes to canary)
+## Step 12 - Test the header-based routing with curl
+
+You can test the header-based routing by sending requests with the specified header.
+With header it always goes to canary:
+
 ```shell
-curl -H "X-test: test" http://localhost:80
+$ kubectl exec -it network-test -c network-test -- sh
+~ # seq 1 100 | xargs -P 10 -I {} bash -c 'curl -s -H "X-Test: test" http://argo-rollouts-service' > pods.txt
+~ # sort pods.txt | uniq -c | sort -rn
+    100 Hello from rollouts-demo-8598f766fd-2cvt4
+~ #
 ```

examples/linkerd-header-based/argo-rollouts-plugin.yaml

@@ -7,4 +7,4 @@ metadata:
 data:
   trafficRouterPlugins: |-
     - name: "argoproj-labs/gatewayAPI"
-      location: "https://github.com/argoproj-labs/rollouts-plugin-trafficrouter-gatewayapi/releases/download/v0.2.0/gateway-api-plugin-linux-arm64"
+      location: "https://github.com/argoproj-labs/rollouts-plugin-trafficrouter-gatewayapi/releases/download/v0.8.0/gatewayapi-plugin-linux-amd64"

examples/linkerd-header-based/kind-cluster.yaml

@@ -1,5 +1,6 @@
 kind: Cluster
 apiVersion: kind.x-k8s.io/v1alpha4
+name: linkerd-header-based
 nodes:
   - role: control-plane
     kubeadmConfigPatches:
@@ -8,10 +9,3 @@ nodes:
         nodeRegistration:
           kubeletExtraArgs:
             node-labels: "ingress-ready=true"
-    extraPortMappings:
-      - containerPort: 80
-        hostPort: 80
-        protocol: TCP
-      - containerPort: 443
-        hostPort: 443
-        protocol: TCP

examples/linkerd-header-based/rollout.yaml

@@ -12,9 +12,10 @@ spec:
       trafficRouting:
         plugins:
           argoproj-labs/gatewayAPI:
-            httpRoute: argo-rollouts-http-route # our created httproute
-            namespace: default # namespace where this rollout resides
-            useHeaderRoutes: true
+            httpRoutes:
+              - name: argo-rollouts-http-route
+                useHeaderRoutes: true
+            namespace: default
         managedRoutes:
           - name: argo-rollouts
       steps:
@@ -30,7 +31,6 @@ spec:
         - pause: {}
         - setHeaderRoute: # remove header route
             name: argo-rollouts
-
   revisionHistoryLimit: 2
   selector:
     matchLabels:
@@ -41,20 +41,22 @@ spec:
         app: rollouts-demo
     spec:
       containers:
-        - name: rollouts-demo
-          image: argoproj/rollouts-demo:red
+        - name: http-echo
+          image: hashicorp/http-echo
+          args:
+          - "-text=Hello from $(POD_NAME)"
           ports:
             - name: http
-              containerPort: 8080
+              containerPort: 5678
               protocol: TCP
           env:
-            - name: APP_VERSION
-              value: "1.0.0"
+          - name: APP_VERSION
+            value: "1.0"
+          - name: POD_NAME
+            valueFrom:
+              fieldRef:
+                fieldPath: metadata.name
           resources:
             requests:
               memory: 32Mi
               cpu: 5m
-
-
-
-