From 486ba5f20b740bb06cd30f66359579077164c27b Mon Sep 17 00:00:00 2001
From: Brent Salisbury <bsalisbu@redhat.com>
Date: Sun, 8 Jun 2025 23:36:22 -0400
Subject: [PATCH] Add /v1/chat/completions to quickstart validation script

Signed-off-by: Brent Salisbury <bsalisbu@redhat.com>
---
 quickstart/README-minikube.md |  4 +--
 quickstart/test-request.sh    | 62 +++++++++++++++++++++++++++++++----
 2 files changed, 57 insertions(+), 9 deletions(-)

diff --git a/quickstart/README-minikube.md b/quickstart/README-minikube.md
index 09e93ab..b63043c 100644
--- a/quickstart/README-minikube.md
+++ b/quickstart/README-minikube.md
@@ -162,8 +162,8 @@ export HF_TOKEN="your-token"
 The inference-gateway serves as the HTTP ingress point for all inference requests in our deployment.
 It’s implemented as a Kubernetes Gateway (`gateway.networking.k8s.io/v1`) using whichever `gatewayClassName` you’ve
 chosen, either `kgateway` or `istio` and sits in front of your inference pods to handle path-based routing, load-balancing,
-retries, and metrics. All calls to `/v1/models` and `/v1/completions` flow through this gateway to the appropriate
-`decode` or `prefill` services.
+retries, and metrics. API requests such as `/v1/models`, `/v1/completions` and `/v1/chat/completions` are exposed via the
+gateway and routed to the appropriate `decode` or `prefill` services.
 
 ```bash
 # -------------------------------------------------------------------------
diff --git a/quickstart/test-request.sh b/quickstart/test-request.sh
index 26f0afc..60a19ce 100755
--- a/quickstart/test-request.sh
+++ b/quickstart/test-request.sh
@@ -122,7 +122,7 @@ validation() {
   echo
 
   # ── 2) POST /v1/completions on decode pod ──────────────────────────────────
-  echo "2 -> Sending a completion request to the decode pod at ${POD_IP}…"
+  echo "2 -> Sending a /v1/completions request to the decode pod at ${POD_IP}…"
   ID=$(gen_id)
   kubectl run --rm -i curl-"$ID" \
     --namespace "$NAMESPACE" \
@@ -136,9 +136,25 @@ validation() {
       }'
   echo
 
-  # 3) GET /v1/models via the gateway
+  # ── 3) POST /v1/chat/completions on decode pod ─────────────────────────────
+  echo "3 -> Sending a /v1/chat/completions request to the decode pod at ${POD_IP}…"
+  ID=$(gen_id)
+  kubectl run --rm -i curl-"$ID" \
+    --namespace "$NAMESPACE" \
+    --image=curlimages/curl --restart=Never -- \
+    curl -sS -X POST http://${POD_IP}:8000/v1/chat/completions \
+      -H 'accept: application/json' \
+      -H 'Content-Type: application/json' \
+      -d '{
+        "model":"'"$MODEL_ID"'",
+        "messages":[{"role":"user","content":"Who are you?"}],
+        "stream":false
+      }'
+  echo
+
+  # ── 4) GET /v1/models via the gateway ───────────────────────────────────────
   GATEWAY_ADDR=$(kubectl get gateway -n "$NAMESPACE" | tail -n1 | awk '{print $3}')
-  echo "3 -> Fetching available models via the gateway at ${GATEWAY_ADDR}…"
+  echo "4 -> Fetching available models via the gateway at ${GATEWAY_ADDR}…"
   ID=$(gen_id)
   GW_JSON=$(kubectl run --rm -i curl-"$ID" \
     --namespace "$NAMESPACE" \
@@ -156,8 +172,8 @@ validation() {
   fi
   echo
 
-  # ── 4) POST /v1/completions via gateway ────────────────────────────────────
-  echo "4 -> Sending a completion request via the gateway at ${GATEWAY_ADDR} with model '${MODEL_ID}'…"
+  # ── 5) POST /v1/completions via gateway ────────────────────────────────────
+  echo "5 -> Sending a /v1/completions request via the gateway at ${GATEWAY_ADDR} with model '${MODEL_ID}'…"
   ID=$(gen_id)
   kubectl run --rm -i curl-"$ID" \
     --namespace "$NAMESPACE" \
@@ -170,6 +186,22 @@ validation() {
         "prompt":"Who are you?"
       }'
   echo
+
+  # ── 6) POST /v1/chat/completions via gateway ──────────────────────────────
+  echo "6 -> Sending a /v1/chat/completions request via the gateway at ${GATEWAY_ADDR} with model '${MODEL_ID}'…"
+  ID=$(gen_id)
+  kubectl run --rm -i curl-"$ID" \
+    --namespace "$NAMESPACE" \
+    --image=curlimages/curl --restart=Never -- \
+    curl -sS -X POST http://${GATEWAY_ADDR}/v1/chat/completions \
+      -H 'accept: application/json' \
+      -H 'Content-Type: application/json' \
+      -d '{
+        "model":"'"$MODEL_ID"'",
+        "messages":[{"role":"user","content":"Who are you?"}],
+        "stream":false
+      }'
+  echo
 }
 
 # ── Minikube gateway validation ───────────────────────────────────────────────
@@ -178,7 +210,7 @@ minikube_validation() {
   echo "Minikube validation: hitting gateway DNS at ${SVC_HOST}"
 
   # 1) GET /v1/models via DNS gateway
-  echo "1 -> GET /v1/models via DNS at ${SVC_HOST}…"
+  echo "1 -> Sending a GET /v1/models via DNS at ${SVC_HOST}…"
   ID=$(gen_id)
   LIST_JSON=$(kubectl run --rm -i curl-"$ID" \
     --namespace "$NAMESPACE" \
@@ -203,7 +235,7 @@ minikube_validation() {
   echo
 
   # 2) POST /v1/completions via DNS gateway
-  echo "2 -> POST /v1/completions via DNS at ${SVC_HOST} with model '${MODEL_ID}'…"
+  echo "2 -> Sending a POST /v1/completions via DNS at ${SVC_HOST} with model '${MODEL_ID}'…"
   ID=$(gen_id)
   kubectl run --rm -i curl-"$ID" \
     --namespace "$NAMESPACE" \
@@ -216,6 +248,22 @@ minikube_validation() {
         "prompt":"You are a helpful AI assistant."
       }'
   echo
+
+  # ── 3) POST /v1/chat/completions via DNS gateway
+  echo "3 -> Sending a POST /v1/chat/completions via DNS at ${SVC_HOST} with model '${MODEL_ID}'…"
+  ID=$(gen_id)
+  kubectl run --rm -i curl-"$ID" \
+    --namespace "$NAMESPACE" \
+    --image=curlimages/curl --restart=Never -- \
+    curl -sS -X POST http://${SVC_HOST}/v1/chat/completions \
+      -H 'accept: application/json' \
+      -H 'Content-Type: application/json' \
+      -d '{
+        "model":"'"$MODEL_ID"'",
+        "messages":[{"role":"user","content":"You are a helpful AI assistant."}],
+        "stream":false
+      }'
+  echo
 }
 
 # ── Main ───────────────────────────────────────────