From f70d4ff62603d6bb7309eb68962d7495116d5884 Mon Sep 17 00:00:00 2001 From: sirius0xdev Date: Wed, 25 Mar 2026 04:17:36 +0000 Subject: [PATCH] fix taints --- apps/base/customer1/news_bot/deepseek-summarizer.yaml | 7 ++++++- modules/a3_nodepool.tf | 3 ++- 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/apps/base/customer1/news_bot/deepseek-summarizer.yaml b/apps/base/customer1/news_bot/deepseek-summarizer.yaml index ec57858..c0668cb 100644 --- a/apps/base/customer1/news_bot/deepseek-summarizer.yaml +++ b/apps/base/customer1/news_bot/deepseek-summarizer.yaml @@ -12,12 +12,17 @@ spec: spec: nodeSelector: accelerator: "h100-single" + cloud.google.com/gke-accelerator: "nvidia-h100-80gb" # 2. ACCEPT THE GPU TAINT tolerations: - key: "nvidia.com/gpu-h100" operator: "Exists" - effect: "NoSchedule" + effect: "NoSchedule" + - key: "nvidia.com/gpu-h100" + operator: "Exists" + effect: "NoSchedule" + volumes: - name: deepseek-storage persistentVolumeClaim: diff --git a/modules/a3_nodepool.tf b/modules/a3_nodepool.tf index 39f636e..f1cf06d 100644 --- a/modules/a3_nodepool.tf +++ b/modules/a3_nodepool.tf @@ -28,7 +28,8 @@ resource "google_container_node_pool" "h100_node_pool" { } labels = { - accelerator = "h100-single" + "accelerator" = "h100-single" + "cloud.google.com/gke-accelerator" = "nvidia-h100-80gb" } disk_size_gb = 200 disk_type = "pd-ssd"