Move the nvidia installer to the beginning.

When the installer runs for the first time, it disables loadpin and restarts the node. So, it is better to run it in the beginning so that we can avoid redoing the later steps. One of the later steps include downloading a tar file and untarring it. Doing that only once saves around 1m30s in test runtime for the gci image.

Move the nvidia installer to the beginning.
4a5badfa · Rohit Agarwal · e72c64b7 · 4a5badfa
Commit 4a5badfa authored Jun 08, 2017 by Rohit Agarwal
Hide whitespace changes
Inline Side-by-side

Showing with 2 additions and 2 deletions

gci-init-gpu.yaml test/e2e_node/jenkins/gci-init-gpu.yaml +2 -2

No files found.
--- a/test/e2e_node/jenkins/gci-init-gpu.yaml
+++ b/test/e2e_node/jenkins/gci-init-gpu.yaml
 #cloud-config

 runcmd:
+  - modprobe configs
+  - docker run -v /dev:/dev -v /home/kubernetes/bin/nvidia:/rootfs/nvidia -v /etc/os-release:/rootfs/etc/os-release -v /proc/sysrq-trigger:/sysrq -e LAKITU_KERNEL_SHA1=26481563cb3788ad254c2bf2126b843c161c7e48 -e BASE_DIR=/rootfs/nvidia --privileged gcr.io/google_containers/cos-nvidia-driver-install@sha256:ad83ede6e0c6d768bf7cf69a7dec972aa5e8f88778142ca46afd3286ad58cfc8
  - mount /tmp /tmp -o remount,exec,suid
  - usermod -a -G docker jenkins
  - mkdir -p /var/lib/kubelet
@@ -15,5 +17,3 @@ runcmd:
  - mount --bind /proc /home/kubernetes/containerized_mounter/rootfs/proc
  - mount --bind /dev /home/kubernetes/containerized_mounter/rootfs/dev
  - rm /tmp/mounter.tar
-  - modprobe configs
-  - docker run -v /dev:/dev -v /home/kubernetes/bin/nvidia:/rootfs/nvidia -v /etc/os-release:/rootfs/etc/os-release -v /proc/sysrq-trigger:/sysrq -e LAKITU_KERNEL_SHA1=26481563cb3788ad254c2bf2126b843c161c7e48 -e BASE_DIR=/rootfs/nvidia --privileged gcr.io/google_containers/cos-nvidia-driver-install@sha256:ad83ede6e0c6d768bf7cf69a7dec972aa5e8f88778142ca46afd3286ad58cfc8