Tools for use with Orchestra 2, an HPC cluster managed by the HMS Resesarch Computing group.
jobinfo.shGet info on SLURM jobs.jupyter_o2Run jupyter over ssh and connect to it in your browser.
A bash script that provides information on your SLURM jobs.
Usage: jobinfo or jobinfo <job id>
Tab autocompletion of the job id is supported.
Source this script in your ~/.bashrc file.
A command line tool to launch and connect to a Jupyter session on O2. (Now available as a package, jupyter-o2)
Usage: jupyter_o2 <USER> <subcommand>
Example: jupyter_o2 js123 notebook
This will follow the procedure described on the
O2 wiki.
jupyter_o2 will launch an X11-enabled SSH session with port forwarding,
start an interactive node running jupyter notebook,
SSH into that interactive node to allow requests to be forwarded,
and finally open the notebook in your browser.
Run jupyter_o2 with ./jupyter_o2 or copy the file into a folder within your $PATH.
- After running jupyter_o2 once, cancel at the PIN entry step using Ctrl-C.
A configuration file should have been created at
~/.jupyter-o2.cfg. Edit this file. - Change
MODULE_LOAD_CALLandSOURCE_JUPYTER_CALLto commands that activate your jupyter environment (more description in the file itself). If one or both of these are not necessary, deleting everything after the=will set it to an empty string. - Choose a
DEFAULT_JP_PORTthat is open on your machine, if the current default (8887) isn't already open. The port can also be specified usingjupyter_o2 -p <port>.
- Pexpect
- dnspython
- pyobjc-framework-Quartz
- pinentry (a command line tool)
jupyter_o2 has been tested with Python versions 2.7 and 3.6, and Pexpect version 4.3. However, other versions may work just as well.
jupyter_o2 has been tested on MacOS. It may work on Linux and it would likely require
both Cygwin and a Cygwin version of Python to work on Windows (for Pexpect and SSH).