Workbench pre-flight check#

Follow the instructions below to install and run the pre-flight script. Run it on each node in the system on which you plan to install Data Science & AI Workbench, to verify that it meets the minimum requirements to install successfully.

To install the package from Git, run the following commands from a terminal on the node you’re checking:

git clone https://github.com/Anaconda-Platform/ae-preflight.git
cd ae-preflight
python setup.py install

To install the package in an existing conda environment, run the following command from a terminal on the node you’re checking:

conda install -c ae5-admin ae_preflight

To run the profiler, run the following command:

ae-preflight

Optionally append the command with any of these options, based on your preferences:

-i, --interface     Interface name (e.g., eth0 or ens3) to check for open ports
-v, --verbose       Increase verbosity of the script
--hostname          Check hostname to ensure that DNS can be resolved for TLD and wildcard DNS

System-specific instructions are also provided below.

The output of running the script is captured in a results.txt file.

To view the results, locate the results.txt file that was created in the directory that you ran the script from.

This file will contain warnings and notify you of any pre-flight checks that may have failed. It will also describe the reason why a test failed, and provide a solution for fixing the issue.

System-specific instructions#

CentOS 7.X

# Install requirements
yum install epel-release -y
yum install git python-psutil -y

# Git clone and install
git clone https://github.com/Anaconda-Platform/ae-preflight.git
cd ae-preflight
python setup.py install

# Run program
ae-preflight

Ubuntu 16.04

# Update apt-caches and install requirements
apt-get update
apt-get install python3-psutil git python3-setuptools -y

# Git clone and install
git clone https://github.com/Anaconda-Platform/ae-preflight.git
cd ae-preflight
python setup.py install

# Run prechecks
ae-preflight

Suse 12

# Install dependencies
zypper install -y git python-psutil

# Git clone and install
git clone https://github.com/Anaconda-Platform/ae-preflight.git
cd ae-preflight
python setup.py install

# Run prechecks
ae-preflight

New conda environments

Installation requirements:

# Install from yum (CentOS and RHEL)
yum install bzip2 -y

# Install from apt (Ubuntu)
apt-get install bzip2 -y

# Install from zypper (Suse)
zypper install -y bzip2

To install and set up a conda environment:

# Get miniconda and install it
curl -O https://repo.anaconda.com/miniconda/Miniconda2-4.6.14-Linux-x86_64.sh
bash Miniconda2-4.6.14-Linux-x86_64.sh  # Accept the license and take the defaults

# Source bashrc to pick up conda paths
source ~/.bashrc

# Create profile, and preflight package
conda create -n python37 python=3.7 psutil -y
conda activate python37
conda install -c ae5-admin ae_preflight -y

# Run prechecks
ae-preflight