Command line interface¶

rjm_config¶

This is an interactive script to set up Globus and funcX on NeSI for use with RJM. It will create a new Globus guest collection (shared directory) on NeSI and check whether the user already has a funcX endpoint running on NeSI. If a funcX endpoint is not already running, one will be created and started on NeSI. If an existing funcX endpoint is found to be running, it will not be restarted. Finally, all configuration values from the above steps will be written to the RJM config file on the local machine and the authentication steps will be run.

While running this script, you will need to enter your NeSI username, password and second factor, your NeSI project code and will need to use a web browser to carry out the Globus authentication as required.

usage: rjm_config [-h] [-l LOGFILE] [-ll {debug,info,warn,error,critical}]
                  [-r] [-w] [-v]

Named Arguments¶

-l, --logfile

logfile. if not specified, all messages will be printed to the terminal.

-ll, --loglevel

Possible choices: debug, info, warn, error, critical

level of log verbosity (setting the level here overrides the config file)

-r, --reauth

Force reauthentication of globus-compute-endpoint

Default: False

-w, --where-config

Print location of the config file and exit

Default: False

-v, --version

show program's version number and exit

rjm_restart¶

This is an interactive script that will restart the funcX endpoint running on NeSI. Sometimes the endpoint gets into a bad state due to network, file system, etc issues on NeSI and this script will attempt to fix it.

While running this script, you will need to enter your NeSI username, password and second factor, your NeSI project code and will need to use a web browser to carry out the Globus authentication as required.

usage: rjm_restart [-h] [-l LOGFILE] [-ll {debug,info,warn,error,critical}]
                   [-v]

Named Arguments¶

-l, --logfile

logfile. if not specified, all messages will be printed to the terminal.

-ll, --loglevel

Possible choices: debug, info, warn, error, critical

level of log verbosity (setting the level here overrides the config file)

-v, --version

show program's version number and exit

rjm_health_check¶

Perform basic checks of the interface to the remote machine

usage: rjm_health_check [-h] [-l LOGFILE]
                        [-ll {debug,info,warn,error,critical}] [-le] [-k] [-r]
                        [-v]

Named Arguments¶

-l, --logfile

logfile. if not specified, all messages will be printed to the terminal.

-ll, --loglevel

Possible choices: debug, info, warn, error, critical

level of log verbosity (default: 'critical')

Default: 'critical'

-le, --logextra

Also log funcx and globus at the chosen loglevel

Default: False

-k, --keep

Keep health check files on remote system, i.e. do not delete them after completing the check (default=False)

Default: False

-r, --retries

Allow retries on function failures

Default: False

-v, --version

show program's version number and exit

rjm_batch_submit¶

Upload files and start jobs

usage: rjm_batch_submit [-h] -f LOCALJOBDIRFILE [-l LOGFILE]
                        [-ll {debug,info,warn,error,critical}] [-le] [--force]
                        [-n] [-c CMD] [-d REMOTEDIR] [-j JOBTYPE] [-m MEM]
                        [-p PROJECTCODE] [-w WALLTIME] [-v]

Named Arguments¶

-f, --localjobdirfile

file that contains the names of the local job directories, one name per line

-l, --logfile

logfile. if not specified, all messages will be printed to the terminal.

-ll, --loglevel

Possible choices: debug, info, warn, error, critical

level of log verbosity (setting the level here overrides the config file)

-le, --logextra

Also log funcx and globus at the chosen loglevel

Default: False

--force

ignore progress from previous runs stored in job directory, i.e. start from scratch

Default: False

-n, --defaultlogname

Use default log name instead of "batch_submit"

Default: False

-c, --cmd