Julia on OLCF JupyterHub Open enclave

This is a step-by-step guide

1. ssh into odo: ssh <username>@login1.odo.olcf.ornl.gov

2. Launch Julia:

    $ module load julia
    $ julia
                _
    _       _ _(_)_     |  Documentation: https://docs.julialang.org
    (_)     | (_) (_)    |
    _ _   _| |_  __ _   |  Type "?" for help, "]?" for Pkg help.
    | | | | | | |/ _` |  |
    | | |_| | | | (_| |  |  Version 1.9.0 (2023-05-07)
    _/ |\__'_|_|_|\__'_|  |  Official https://julialang.org/ release
    |__/                   |
   

3. Create a JupyterHub kernel with IJulia

    julia> using IJulia
   
    julia> installkernel("Julia-24threads", "--project=@.", env=Dict("LD_LIBRARY_PATH" => "", "JULIA_NUM_THREADS" => "24"))
   

4. The above would create a config file to launch the Julia kernel next time we open Jupyter OLCF open

    $ cat /ccsopen/home/wgodoy/.local/share/jupyter/kernels/julia-24threads_-1.10/kernel.json 
    {
      "display_name": "Julia-24threads) 1.10.4",
      "argv": [
      "/autofs/nccsopen-svm1_sw/odo/julia-1.10.4/bin/julia",
      "-i",
      "--color=yes",
      "/ccsopen/home/wgodoy/.julia/packages/IJulia/Vo51o/src/kernel.jl",
      "{connection_file}"
      ],
      "language": "julia",
      "env": {
         "LD_LIBRARY_PATH": "",
         "JULIA_NUM_THREADS": "24"
      },
      "interrupt_mode": "signal"
    }
   

5. Log into Jupyter OLCF open and select the kernel Julia-24threads) 1.10.4 to start a new notebook. Example provided in GrayScott.jl/Notebooks/Plot2D.jl/src/Julia-reading.ipynb. Package dependencies need to be precompiled on the first run, instantiate: GrayScott-JACC/Notebooks/Plot2D.jl/Project.toml.

Julia’s own Pluto.jl

Pluto.jl is Julia’s own alternative to using Notebooks. It takes advantage of the fact that Julia manages its own packaging ecosystem. It’s reactive by nature, so every update in the Pluto notebook would trigger the package manager and there is no need to pre-define an environment.

To trigger Pluto.jl:

$ julia
julia> using Pluto
julia> Pluto.run()