refactor(deploy): use docker run instead of helm test, metrics dashboard scripts

scripts/run_on_tester

@@ -22,4 +22,5 @@ ENV_VARS="
 # Format the environment variables for the SSH command
 ENV_EXPORTS=$(printf 'export %s; ' $ENV_VARS)
 
+echo "Running on tester at IP: $SPOT_IP"
 ssh -o ControlMaster=auto -o ControlPath=~/.ssh_mux_%h_%p_%r -o ControlPersist=30s -o TCPKeepAlive=no -o ServerAliveCountMax=5 -o ServerAliveInterval=30 -o StrictHostKeyChecking=no -i "$SPOT_KEY" ubuntu@"$SPOT_IP" "$ENV_EXPORTS $@"

spartan/aztec-network/values.yaml

@@ -1,14 +1,8 @@
-test: "sample"
-scenario: "default"
-
 network:
   public: false
   enableBots: true
 
 images:
-  test:
-    image: aztecprotocol/end-to-end
-    pullPolicy: IfNotPresent
   aztec:
     image: aztecprotocol/aztec
     pullPolicy: IfNotPresent
@@ -182,4 +176,4 @@ metrics:
     periodSeconds: 10
     timeoutSeconds: 5
     successThreshold: 1
-    failureThreshold: 3
+    failureThreshold: 3

spartan/scripts/forward_metrics_dashboard.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+echo "Run this script after the metrics pods all come up. After running this script, port forward 5601 locally"
+echo "Then navigate to https://localhost:5601 and login with user elastic password admin"
+kubectl port-forward -n metrics service/metrics-eck-kibana-kb-http 5601:5601

spartan/scripts/setup_local_k8s.sh

@@ -30,4 +30,8 @@ if ! command -v helm &> /dev/null; then
   rm get_helm.sh
 fi
 
-kind create cluster
+if kind get clusters | grep -q "^kind$"; then
+  echo "Cluster 'kind' already exists. Skipping creation."
+else
+  kind create cluster
+fi

spartan/scripts/setup_metrics_dashboard.sh

@@ -0,0 +1,10 @@
+#!/bin/bash
+
+helm repo add open-telemetry https://open-telemetry.github.io/opentelemetry-helm-charts
+helm repo add elastic https://helm.elastic.co
+helm repo update
+helm dependency build "$(git rev-parse --show-toplevel)/spartan/metrics/"
+kubectl create -f https://download.elastic.co/downloads/eck/2.14.0/crds.yaml 2>/dev/null
+kubectl create secret generic elasticsearch-es-elastic-user --from-literal=elastic=admin -n metrics
+kubectl apply -f https://download.elastic.co/downloads/eck/2.14.0/operator.yaml
+helm upgrade --install metrics "$(git rev-parse --show-toplevel)/spartan/metrics/" --wait --namespace metrics --create-namespace --wait-for-jobs=true --timeout=30m

yarn-project/end-to-end/Earthfile

@@ -299,9 +299,9 @@ e2e-cli-wallet:
 network-smoke:
   ARG values_file
   LOCALLY
-  RUN NAMESPACE=smoke FRESH_INSTALL=true FORCE_BUILD=$force_build VALUES_FILE=${values_file:-default.yaml} ./scripts/network_test.sh ./src/spartan/smoke.test.ts
+  RUN NAMESPACE=smoke FRESH_INSTALL=true VALUES_FILE=${values_file:-default.yaml} ./scripts/network_test.sh ./src/spartan/smoke.test.ts
 
 network-transfer:
   ARG values_file
   LOCALLY
-  RUN NAMESPACE=transfer FRESH_INSTALL=true FORCE_BUILD=$force_build VALUES_FILE=${values_file:-default.yaml} ./scripts/network_test.sh ./src/spartan/transfer.test.ts
+  RUN NAMESPACE=transfer FRESH_INSTALL=true VALUES_FILE=${values_file:-default.yaml} ./scripts/network_test.sh ./src/spartan/transfer.test.ts

yarn-project/end-to-end/scripts/network_test.sh

@@ -43,18 +43,45 @@ if [ "$FRESH_INSTALL" = "true" ]; then
   kubectl delete namespace "$NAMESPACE" --ignore-not-found=true --wait=true --now --timeout=10m
 fi
 
+function show_status_until_pxe_ready() {
+  set +x # don't spam with our commands
+  sleep 15 # let helm upgrade start
+  for i in {1..100} ; do
+    if kubectl wait pod -l app==pxe --for=condition=Ready -n "transfer" --timeout=20s >/dev/null 2>/dev/null ; then
+      break # we are up, stop showing status
+    fi
+    # show startup status
+    kubectl get pods -n "$NAMESPACE"
+  done
+}
+
+show_status_until_pxe_ready &
+
 # Install the Helm chart
-helm install spartan "$(git rev-parse --show-toplevel)/spartan/aztec-network/" \
+helm upgrade --install spartan "$(git rev-parse --show-toplevel)/spartan/aztec-network/" \
       --namespace "$NAMESPACE" \
       --create-namespace \
       --values "$(git rev-parse --show-toplevel)/spartan/aztec-network/values/$VALUES_FILE" \
-      --set images.test.image="aztecprotocol/end-to-end:$AZTEC_DOCKER_TAG" \
       --set images.aztec.image="aztecprotocol/aztec:$AZTEC_DOCKER_TAG" \
-      --set test="$TEST" \
+      --set ingress.enabled=true \
       --wait \
       --wait-for-jobs=true \
       --timeout=30m
 
 kubectl wait pod -l app==pxe --for=condition=Ready -n "$NAMESPACE" --timeout=10m
 
-helm test spartan --namespace "$NAMESPACE" --timeout 30m
+function forward_pxe_k8s_port() {
+  # NOTE we fail silently, and work in the background
+  kubectl port-forward --namespace transfer svc/spartan-aztec-network-pxe 9082:8080 2>/dev/null >/dev/null || true
+}
+# tunnel in to get access directly to our PXE service in k8s
+(kubectl port-forward --namespace transfer svc/spartan-aztec-network-pxe 9082:8080 2>/dev/null >/dev/null || true) &
+
+# run our test in the host network namespace (so we can access the above with localhost)
+docker run --rm --network=host \
+  -e SCENARIO=default \
+  -e PXE_URL=http://localhost:9082 \
+  -e DEBUG="aztec:*" \
+  -e LOG_LEVEL=debug \
+  -e LOG_JSON=1 \
+  aztecprotocol/end-to-end:$AZTEC_DOCKER_TAG $TEST