From a053e925ae0a6d52ce8403a6392fd1fe90a4503f Mon Sep 17 00:00:00 2001 From: Chirag Jain Date: Tue, 16 Jan 2024 19:29:04 +0530 Subject: [PATCH] EKS GPUs - Disable GSP from userData script (#176) --- charts/tfy-karpenter-config/Chart.yaml | 2 +- charts/tfy-karpenter-config/values.yaml | 11 +++++++++++ 2 files changed, 12 insertions(+), 1 deletion(-) diff --git a/charts/tfy-karpenter-config/Chart.yaml b/charts/tfy-karpenter-config/Chart.yaml index 55a7e11b9..58a41c3db 100644 --- a/charts/tfy-karpenter-config/Chart.yaml +++ b/charts/tfy-karpenter-config/Chart.yaml @@ -2,4 +2,4 @@ apiVersion: v2 name: tfy-karpenter-config description: "ArgoCD Applications for karpenter config" type: application -version: 0.1.18-rc.6 +version: 0.1.18-rc.7 diff --git a/charts/tfy-karpenter-config/values.yaml b/charts/tfy-karpenter-config/values.yaml index 4b07ba276..8a833c067 100644 --- a/charts/tfy-karpenter-config/values.yaml +++ b/charts/tfy-karpenter-config/values.yaml @@ -227,6 +227,17 @@ karpenter: echo "CONTAINERD_VERSION is empty or not within the specified range." fi + rmmod nvidia_drm + rmmod nvidia_modeset + rmmod nvidia_uvm + rmmod nvidia + echo "Writing NVreg_EnableGpuFirmware=0 to /etc/modprobe.d/nvidia.conf" + echo "options nvidia NVreg_EnableGpuFirmware=0" | tee --append /etc/modprobe.d/nvidia.conf + echo "Writing NVreg_EnableGpuFirmware=0 to /etc/modprobe.d/nvidia-gsp.conf" + echo "options nvidia NVreg_EnableGpuFirmware=0" | tee --append /etc/modprobe.d/nvidia-gsp.conf + echo "Running dracut" + dracut -f + ## Default Inferentia Node Template Config inferentiaDefaultNodeTemplate: ## @param karpenter.inferentiaDefaultNodeTemplate.rootVolumeSize Size for the root volume attached to node