You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mxnet.apache.org by GitBox <gi...@apache.org> on 2021/03/30 06:44:19 UTC

[GitHub] [incubator-mxnet-ci] marcoabreu commented on a change in pull request #37: Add new config for mxnetlinux-gpu-g4 nodes.

marcoabreu commented on a change in pull request #37:
URL: https://github.com/apache/incubator-mxnet-ci/pull/37#discussion_r603822612



##########
File path: tools/ami-creator/userdata/mxnetlinux_gpu_ubuntu_2004.txt
##########
@@ -0,0 +1,81 @@
+#cloud-config
+
+apt_reboot_if_required: false
+package_update: true
+package_upgrade: true
+apt:
+  sources:
+    nvidia-cuda-drivers.list:
+      source: |
+        deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64 /
+      keyid: 7FA2AF80
+    docker.list:
+      source: deb [arch=amd64] https://download.docker.com/linux/ubuntu $RELEASE stable
+      keyid: 0EBFCD88
+    nvidia-container-runtime.list:
+      source: |
+        deb https://nvidia.github.io/libnvidia-container/ubuntu20.04/$(ARCH) /
+        deb https://nvidia.github.io/nvidia-container-runtime/ubuntu20.04/$(ARCH) /
+        deb https://nvidia.github.io/nvidia-docker/ubuntu20.04/$(ARCH) /
+      keyid: F796ECB0
+
+apt_get_command: ["apt-get", "--option=Dpkg::Options::=--force-confold", "--option=Dpkg::options::=--force-unsafe-io", "--assume-yes", "--quiet", "--install-recommends"]
+
+packages:
+  - wget
+  - curl
+  - openjdk-8-jre
+  - git
+  - python3
+  - python3-pip
+  - cuda-drivers
+  - apt-transport-https
+  - ca-certificates
+  - software-properties-common
+  - docker-ce
+  - nvidia-container-toolkit
+  - nvidia-container-runtime
+
+write_files:
+  - path: /etc/cron.d/jenkins-start-slave
+    content: |
+      @reboot jenkins_slave /home/jenkins_slave/scripts/launch-autoconnect.sh
+  - path: /home/jenkins_slave/scripts/launch-autoconnect.sh
+    content: |
+      #!/bin/sh
+      set -ex
+      sleep 300

Review comment:
       Are you really sure about that? The entire auto scaling is built on the premise that the instance quickly starts up. Just introducing this sleep would be really harmful for the overall performance. Also, there are no package installations happening when the slave starts - that kind of stuff happens during AMI creation.




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org