It suggested this page develop as problems arise and are recognised.

Diagnostics

Check what is running on Archer2

squeue -u $USER -l

Check what is running on PUMA

cylc gscan &

right click on suites showing an error - red square - to open that suite in gcylc.
ungroup the tasks (RUN_MAIN and HPC_SERIAL usually)
right click on  a task that failed to see the log.err and log.out

if nothing is running, has it finished?

• all suites concluded, no pending SLURM job
• study json with _final in path name and with costd with in it.
• last jobOutput/RE file concludes with statement about reaching: ** its target ** maximum number of models allowed to be run (maxfun in the suite's json)

was there a problem with environment or the suite's json?

On ARCHER2

• was ". ~/setup_optclim.sh" run and the expected python environment set in the PATH ("which python")

• Did runAlgorthm.py work?

• check trace from initial run the latest suitedir/jobOutput/RE... file

• check the expected interface directories exist with *** runParams.txt set up? *** observables,json hold ing simulate dobservables ( Are there error messages in the jobOutput/PP... files?)

• check the latest suitedir/jobOutput/RE... file