Node#execute runs but node cannot be accessed from command-line

Mason,

Here is a longer explanation of how to properly store your keys and SSH configuration in Jupyter Hub:

The keys in /home/fabric/.ssh/ are not persistent across the container boots. It’s user’s responsibility to upload their keys at a location under /home/fabric/work directory, which is persistent. They can then point FABLib to use that as below:

By default a rc file with the required environment variables is created at /home/fabric/work/fabric_config/fabric_rc

export FABRIC_BASTION_KEY_LOCATION=/home/fabric/work/fabric_config/fabric-bastion
export FABRIC_SLICE_PRIVATE_KEY_FILE=/home/fabric/work/fabric_config/id_rsa
export FABRIC_SLICE_PUBLIC_KEY_FILE=/home/fabric/work/fabric_config/id_rsa.pub

Workflow for SSH:

1. Update /home/fabric/work/fabric_config/ssh_config to reflect the correct keys and bastion user name
   SSH to the VMs using the following command

ssh -F ~/work/fabric_config/ssh_config -i ~/work/fabric_config/id_rsa <username>@<mgmt-ip>

Workflow for Notebooks:

1. Update the environment variables in /home/fabric/work/fabric_config/fabric_rc
2. Import the environment variables in the notebooks as below

fabric_rc_location=os.environ[‘HOME’]+”/work/fabric_config/fabric_rc”
   if os.path.exists(fabric_rc_location):
      with open(fabric_rc_location, ‘r’) as f:
         for line in f:
            if line.startswith(‘export’):
               os.environ[line.split(‘=’)[0].split(‘export’)[1].strip()] = line.split(‘=’)[1].strip()

ssh_key_file_priv=os.environ[“FABRIC_SLICE_PRIVATE_KEY_FILE”]
ssh_key_file_pub=os.environ[“FABRIC_SLICE_PUBLIC_KEY_FILE”]

ssh_key_pub = None
with open (ssh_key_file_pub, “r”) as myfile:
   ssh_key_pub=myfile.read()
   ssh_key_pub=ssh_key_pub.strip()