Create an accelerated image

Go to the instance purchase page and select the Custom Launch tab. Configure the instance to meet the following requirements. For more information, see Custom purchase of an instance.

• Billing Method: Select Pay-as-you-go.

• Instance Type: Select an instance type with the x86_64 architecture.

• Image: Select an image from the CentOS 7 series.

• Public IP Address: Select Assign Public IPv4 Address.

• Security Group: Select the security group that your cluster uses.

Click Create Order to create the instance.

After the instance status changes to Running, connect to the ECS instance that you created. For more information, see Connect to a Linux instance by using Workbench.

Download and install the required Python environments.

1. In the root directory, create an installation script named install_python.sh.

   #!/bin/bash
   
   PYTHON_INSTALL_ROOT=/usr/local/py3.9env
   SERVICE_INSTALL_ROOT=/usr/local/ehpc_service
   
   mkdir $PYTHON_INSTALL_ROOT
   
   set -e
   
   arch=`arch`
   srcEnvPath=$1  # Pass this as an argument when calling the script
   #py3=$2  # Pass this as an argument (0 or 1) to decide which Python version to install
   
   pythonenv="python3.9env"
   echo "Downloading and extracting Python 3.9.10 for ${arch}..."
   wget -c -nv ${srcEnvPath}/python3.9.10.$arch.tgz -O - | tar -xz -C /usr/local/py3.9env/
   echo "Downloading and extracting ${pythonenv} for ${arch}..."
   curl -s -O ${srcEnvPath}/${pythonenv}.$arch.tgz
   if [ -d /usr/local/py3.9env/${pythonenv} ]; then
     rm -rf /usr/local/py3.9env/${pythonenv}
   fi
   tar -xzf ${pythonenv}.$arch.tgz -C /usr/local/py3.9env/
   
   echo "Downloading and extracting Python 2.7 environment..."
   curl -s -O ${srcEnvPath}/centos/7.2/python2.7env.tgz
   if [ -d /usr/local/python2.7env ]; then
     rm -rf /usr/local/python2.7env
   fi
   tar -xzf python2.7env.tgz -C /usr/local/

2. Grant execute permissions to the script.

   chmod +x install_python.sh

3. Run the script to download and install the environments.

   bash /root/install_python.sh http://public-ehs.vpc100-oss-cn-hangzhou.aliyuncs.com/packages/env

4. Verify that the py3.9env and python2.7env directories were created in the /usr/local/ directory to confirm a successful installation.

Download and install Node.js and other dependency packages.

1. In the root directory, create a download and installation script named install_lib.sh.

   #!/bin/bash
   
   set -e
   arch=`arch`  # Pass this as an argument when calling the script
   srcEnvPath=$1  # Pass this as an argument when calling the script
   
   # echo "Installing Node.js for aarch64..."
   # curl -s -O $srcEnvPath/node-v16.17.1-linux-arm64.tar.xz
   # tar -xf node-v16.17.1-linux-arm64.tar.xz
   # mv node-v16.17.1-linux-arm64 /usr/local/node-v16.17.1-linux-arm64
   # ln -sfT /usr/local/node-v16.17.1-linux-arm64/bin/node /usr/local/bin/node
   # ln -sfT /usr/local/node-v16.17.1-linux-arm64/bin/npm /usr/local/bin/npm
   
   echo "Installing Node.js for x86_64..."
   curl -s -O $srcEnvPath/node-v6.11.1-linux-x64.tar.gz
   tar -xzf node-v6.11.1-linux-x64.tar.gz
   mv node-v6.11.1-linux-x64 /usr/local/node-v6.11.1
   ln -sfT /usr/local/node-v6.11.1/bin/node /usr/local/bin/node
   ln -sfT /usr/local/node-v6.11.1/bin/npm /usr/local/bin/npm
   
   echo "Installing additional libraries..."
   yum clean all && yum makecache
   yum install -y perf postgresql-contrib hwloc-devel jsoncpp-devel libcurl-devel nfs-utils cifs-utils environment-modules unzip jq
   
   echo "Appending module path..."
   echo -e "\nexport MODULEPATH=/opt/ehpcmodulefiles" >> /etc/bashrc
   
   kernel_version="$(uname -r)"
   kernel_devel_pkg="kernel-devel-${kernel_version}"
   base_os="7"
   kernel_devel_remote_repo="http://mirrors.cloud.aliyuncs.com/opsx/ecs/linux/rpm/driver/${base_os}/x86_64"
   
   rpm -q "${kernel_devel_pkg}" || yum install -y "${kernel_devel_pkg}"
   
   
   echo "Installing xdragon-health-protector..."
   xdragon_pkg="xdragon-health-protector-vm-1.0.x86_64.rpm"
   curl -s -O $srcEnvPath/${xdragon_pkg}
   yum install -y ${xdragon_pkg}
   systemctl enable xdragon-health-protector
   systemctl restart xdragon-health-protector
   systemctl status xdragon-health-protector

2. Grant execute permissions to the script.

   chmod +x install_lib.sh

3. Run the installation script. A running status indicates a successful installation.

   bash /root/install_lib.sh http://public-ehs.vpc100-oss-cn-hangzhou.aliyuncs.com/packages/env

   

Go to the ECS console - Instances page and click the ID of the instance you created to go to the Instance Details page.

Click Create a custom image to generate the accelerated image.

Log on to the E-HPC console. Click the target Cluster Name to go to the Cluster Details page.

In the left-side navigation pane, choose Nodes and Queues > Queue. In the Operation column of the target queue, click Edit.

On the Edit Queue page, configure the parameters for the Instance type Group. For more information, see Configure an auto-scaled node.

1. Click Add an instance specification and select the instance types for the nodes.

2. Click Custom image and select the accelerated image that you created.

   

3. Click OK.

Click Save to apply the auto scaling configuration.