From bd760287a8af42ccd6c39479bc78b1a725e1b7ae Mon Sep 17 00:00:00 2001
From: sirius0xdev <sirius0xdev@users.noreply.github.com>
Date: Wed, 29 Apr 2026 16:01:03 +0000
Subject: [PATCH] fix(rtx6000): enable scale-to-zero with 20min cooldown +
 wake-up route

- Set replicas.min: 0 for scale-to-zero when idle
- Add config.cooldownPeriod: 1200 to KEDA HTTP add-on HelmRelease
  (20 min buffer before scaling down from 1 replica)
- Add external HTTPRoute at brain.siriusdevops.com for manual wake-up
  via curl from Hermes/Telegram
---
 .../base/keda/release-http-add-on.yaml        |  3 +++
 .../gpus/base/keda-gpu-scaling/keda-vllm.yaml |  2 +-
 .../base/keda-gpu-scaling/kustomization.yaml  |  1 +
 .../base/keda-gpu-scaling/vllm-route.yaml     | 19 +++++++++++++++++++
 4 files changed, 24 insertions(+), 1 deletion(-)
 create mode 100644 infrastructure/gpus/base/keda-gpu-scaling/vllm-route.yaml

diff --git a/infrastructure/controllers/base/keda/release-http-add-on.yaml b/infrastructure/controllers/base/keda/release-http-add-on.yaml
index a57959a..3f01b7c 100644
--- a/infrastructure/controllers/base/keda/release-http-add-on.yaml
+++ b/infrastructure/controllers/base/keda/release-http-add-on.yaml
@@ -19,3 +19,6 @@ spec:
     crds: CreateReplace
   dependsOn:
     - name: keda
+  values:
+    config:
+      cooldownPeriod: 1200
diff --git a/infrastructure/gpus/base/keda-gpu-scaling/keda-vllm.yaml b/infrastructure/gpus/base/keda-gpu-scaling/keda-vllm.yaml
index 34dea56..ef7278d 100644
--- a/infrastructure/gpus/base/keda-gpu-scaling/keda-vllm.yaml
+++ b/infrastructure/gpus/base/keda-gpu-scaling/keda-vllm.yaml
@@ -13,6 +13,6 @@ spec:
     service: rtx6000-brain-service
     port: 8000
   replicas:
-    min: 1
+    min: 0
     max: 1
  
diff --git a/infrastructure/gpus/base/keda-gpu-scaling/kustomization.yaml b/infrastructure/gpus/base/keda-gpu-scaling/kustomization.yaml
index b1a2fcc..abd516b 100644
--- a/infrastructure/gpus/base/keda-gpu-scaling/kustomization.yaml
+++ b/infrastructure/gpus/base/keda-gpu-scaling/kustomization.yaml
@@ -3,4 +3,5 @@ apiVersion: kustomize.config.k8s.io/v1beta1
 kind: Kustomization
 resources:
   - keda-vllm.yaml
+  - vllm-route.yaml
 
diff --git a/infrastructure/gpus/base/keda-gpu-scaling/vllm-route.yaml b/infrastructure/gpus/base/keda-gpu-scaling/vllm-route.yaml
new file mode 100644
index 0000000..a4faefb
--- /dev/null
+++ b/infrastructure/gpus/base/keda-gpu-scaling/vllm-route.yaml
@@ -0,0 +1,19 @@
+
+apiVersion: gateway.networking.k8s.io/v1
+kind: HTTPRoute
+metadata:
+  name: rtx6000-brain-route
+  namespace: customer1
+spec:
+  parentRefs:
+  - name: external-http-gateway
+  hostnames:
+  - brain.siriusdevops.com
+  rules:
+  - matches:
+    - path:
+        type: PathPrefix
+        value: /
+    backendRefs:
+    - name: rtx6000-brain-service
+      port: 8000