Jupiter interactive app LHA

Hello everyone! I am a newbie here. I’ve managed to install Open OnDemand with Dex LDAP authentication, and now I’m trying to run Jupyter using this tutorial:


Some details:

  • Resource manager: LinuxHost Adapter
  • Host based authentication working between OOD node and target host
  • Singularity Jupyter image tested working on target host

I’ve reached the end of the tutorial, but Jupyter doesn’t appear to be running.

This is what I do:

But nothing more, no Jupyter web interface is available. On the compute node, I don’t see any ports open, nevertheless, I see OOD node logging in.
No relevant information on /var/log/ondemand-nginx/user/error.log either.
I don’t know how to continue from here, can anyone give me some hint?

Look at the output.log file in the $HOME/ondemand/data/sys/dashboard/batch_connect/sys/bc_jupyter directory

Thanks for your response Liu. I don’t have that file/directory:

$ ls $HOME/ondemand/data/sys/dashboard/batch_connect/sys/jupyter/
context.json output

$ ls $HOME/ondemand/data/sys/dashboard/batch_connect/sys/jupyter/output/
3ea6ae52-e19e-4058-8f9a-4643d043a359 6198c968-4350-416c-bfac-c857d1f9e8b7 da7a315e-2251-433d-b312-c3cde94db838
588be745-be2f-4427-9a0e-23052b9da49f d2c16b8f-1a6a-43d1-b8c5-cab46ad3dc69

$ ls $HOME/ondemand/data/sys/dashboard/batch_connect/sys/jupyter/output/3ea6ae52-e19e-4058-8f9a-4643d043a359/
after.sh before.sh job_script_content.sh job_script_options.json script.sh user_defined_context.json

Cluster file: /etc/ood/config/clusters.d/jupyter.yml

Thanks for your help

View the contents of the following files
├── form.yml
├── manifest.yml
├── submit.yml.erb
├── template
│ ├── before.sh.erb
│ └── script.sh.erb
└── view.html.erb

The files are based on GitHub - OSC/bc_example_jupyter: Batch Connect - Example Jupyter Notebook Server
Only form.yml has a change:
cluster: “jupyter” (this is correct)

manifest.yml, submit.yml.erb, before.sh.erb, script.sh.erb, view.html.erb without changes

What is the file content of submit.yml.erb and script.sh.erb?


# Job submission configuration file

# Configure the content of the job script for the batch job here
# @see http://www.rubydoc.info/gems/ood_core/OodCore/BatchConnect/Template
  # We use the basic web server template for generating the job script
  # @note Do not change this unless you know what you are doing!
  template: "basic"

  # You can override the command used to query the hostname of the compute node
  # here
  # @note It is **highly** recommended this be set in the global cluster
  #   configuration file so that all apps can take advantage of it by default
  #set_host: "host=$(hostname -A | awk '{print $2}')"

# Configure the job script submission parameters for the batch job here
# @see http://www.rubydoc.info/gems/ood_core/OodCore/Job/Script
#  queue_name: "queue1"
#  accounting_id: "account1"
#  email_on_started: true
#  native: # ... array of command line arguments ...

#!/usr/bin/env bash

# Benchmark info
echo "TIMING - Starting main script at: $(date)"

# Set working directory to home directory
cd "${HOME}"

# Start Jupyter Notebook Server

<%- unless context.modules.blank? -%>
# Purge the module environment to avoid conflicts
module purge

# Load the require modules
module load <%= context.modules %>

# List loaded modules
module list
<%- end -%>

# Benchmark info
echo "TIMING - Starting jupyter at: $(date)"

# Launch the Jupyter Notebook Server
set -x
jupyter notebook --config="${CONFIG_FILE}" <%= context.extra_jupyter_args %>

My file contents

  template: "basic"
    - "-N"
    - "<%= bc_num_slots.blank? ? 1 : bc_num_slots.to_i %>"
cluster: "linux"
  modules: "python/3.8"
  extra_jupyter_args: ""
  - modules
  - extra_jupyter_args
  - bc_account
  - bc_queue
  - bc_num_hours
  - bc_num_slots
  - bc_email_on_started