Python Virtual Environment Setup

• Download and run the script on the x86 server.

  1 2	wget https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/Anaconda3-2.1.0-Linux-x86_64.sh sh Anaconda3-2.1.0-Linux-x86_64.sh

• Download and run the script on the Arm server.

  1 2	wget https://github.com/Archiconda/build-tools/releases/download/0.2.3/Archiconda3-0.2.3-Linux-aarch64.sh sh Archiconda3-0.2.3-Linux-aarch64.sh

1. Execute the Archiconda3-0.2.3-Linux-aarch64.sh file.

   sh Archiconda3-0.2.3-Linux-aarch64.sh

   

2. Enter yes to accept the agreement.

   

3. Enter the Anaconda installation directory or press Enter to install Anaconda in the default directory /root/archiconda3. In this example, the installation directory defaults to /root/archiconda3.

   

4. Enter yes to create the environment variable.

   

5. Update the environment variable.

   1	source .bash_profile

1. Create a Python virtual environment on the x86 server.

   1	conda create -n mlpy_env --copy -y -q python=2

2. Create a Python virtual environment on the Arm server.

   1	conda create -n mlpy_env_aarch64 --copy -y -q python=2

After the creation is completed, a virtual environment folder is generated in the envs directory of the Anaconda installation directory. For example, if Anaconda is installed in the default directory, a virtual environment folder is generated in the /root/archiconda3/envs directory.

1. Copy the mlpy_env and mlpy_env_aarch64 directories in the envs directory of the Archiconda installation directory on both the x86 and Arm servers to the same directory, for example, /home, of any server.
2. Compress the mlpy_env and mlpy_env_aarch64 directories into a ZIP package.

   1 2	cd /home zip -r mlpy_env.zip mlpy_env mlpy_env_aarch64

   

   The mlpy_env.zip package is the Python package that can be run on both x86 and Arm servers.

3. Upload mlpy_env.zip to the /opt directory on the task submitting host in the cluster.