From 4da2a4c81384f9f17bed69982f5a82200a741767 Mon Sep 17 00:00:00 2001
From: lingyuzeng <pylyzeng@gmail.com>
Date: Thu, 18 Jul 2024 10:43:06 +0800
Subject: [PATCH] add system package

---
 spawnerdockerfile/Dockerfile.ngc | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/spawnerdockerfile/Dockerfile.ngc b/spawnerdockerfile/Dockerfile.ngc
index 857b36c..d474ee7 100644
--- a/spawnerdockerfile/Dockerfile.ngc
+++ b/spawnerdockerfile/Dockerfile.ngc
@@ -22,14 +22,16 @@ RUN <<EOT
 #!/bin/bash
 # SYSTEM_NAME=$(lsb_release -cs) # 查看发行版本
 # Pre-build **latest** DeepSpeed, so it would be ready for testing (otherwise, the 1st deepspeed test will timeout)
-python3 -m pip uninstall -y deepspeed
 # This has to be run (again) inside the GPU VMs running the tests.
 # The installation works here, but some tests fail, if we do not pre-build deepspeed again in the VMs running the tests.
 # TODO: Find out why test fail. install deepspeed
 # DS_BUILD_CPU_ADAM=${DS_BUILD_CPU_ADAM} DS_BUILD_FUSED_ADAM={DS_BUILD_FUSED_ADAM} python3 -m pip install "deepspeed<=0.14.0" --global-option="build_ext" --global-option="-j8" --no-cache -v --disable-pip-version-check 2>&1
 # from https://github.com/huggingface/transformers/blob/main/docker/transformers-pytorch-deepspeed-latest-gpu/Dockerfile install deepspeed fail 
 # reference deepspeed install from https://github.com/microsoft/DeepSpeed/blob/master/docker/Dockerfile
-# install deepspeed prepare
+# install Mellanox OFED prepare
+apt-get update
+apt install libnvidia-compute-535
+apt-get install -y pciutils tk kmod libusb-1.0-0 tcl chrpath libpci3 bison lsof graphviz ethtool swig udev libltdl-dev libelf1 libmnl0 debhelper flex libfuse2
 # install Mellanox OFED
 mkdir -p ${STAGE_DIR}
 wget -q -O - http://www.mellanox.com/downloads/ofed/MLNX_OFED-${MLNX_OFED_VERSION}/MLNX_OFED_LINUX-${MLNX_OFED_VERSION}-ubuntu22.04-x86_64.tgz | tar xzf - -C ${STAGE_DIR}