Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Errors in scripts_regression_tests.py

inos@bas_ac_uk

Ingrid Cnossen
Member
Hello,I am trying to port CESM 2.1.0 to a new machine called Archer. I have previously ported CESM 1.0.5 to this machine, several years ago, but things have changed and I am running into problems with this new port. I have done my best to define customized config_machines.xml, config_compilers.xml, and config_batch.xml files for Archer, which I have attached for information. However, when I run scripts_regression_tests.py I get several failures. A lot of them have the following form:Check case OK
Submitting job script qsub -q default -l walltime=0:05:00 -A n02-bas -v ARGS_FOR_SCRIPT='--skip-preview-namelist' .case.test
submit_jobs case.test
Submit job case.test
ERROR: Command: 'qsub -q default -l walltime=0:05:00 -A n02-bas -v ARGS_FOR_SCRIPT='--skip-preview-namelist' .case.test' failed with error 'qsub: illegal -N value
usage: qsub [-a date_time] [-A account_string] [-c interval]
        [-C directive_prefix] [-e path] [-f ] [-h ] [-I [-X]] [-j oe|eo] [-J X-Y[:Z]]
        [-k o|e|oe] [-l resource_list] [-m mail_options] [-M user_list]
        [-N jobname] [-o path] [-p priority] [-q queue] [-r y|n]
        [-S path] [-u user_list] [-W otherattributes=value...]
        [-v variable_list] [-V ] [-z] [script | -- command [arg1 ...]]
qsub --version' from dir '/fs2/n02/n02/inos/modeldata/CESM2.1.0/scripts_regression_test.20190108_095040/SMS.T42_T42.S.archer_intel.fake_testing_only_20190108_112231'The full TestStatus.log file for one of the tests is attached as an example. The error message suggests to me that there is a problem with the format of the job name, and Google tells me that PBS may have a problem with a job name starting with a ".", so perhaps the fact that the job is called ".case.test" results in the error. This should be easy to check, but I am not sure how all the scripts work together and therefore I don't know where/how I can change the job name to test if this is the issue. I am also a bit surprised that no one else has (apparently) stumbled across this, which makes me wonder if there's actually something else I'm doing wrong, but I'm not sure what... Any advice would be much appreciated!Thanks,Ingrid
 

inos@bas_ac_uk

Ingrid Cnossen
Member
I have done some further testing, and I no longer think that the issue is with the job submission script being called ".case.test", but rather with the length of the job name specified within ".case.test". Apparently PBS does not like long job names and accepts only job names with 15 characters or less. The test names tend to be much longer than this. There must be a way to truncate these; I guess I might be able to specify something in my ~/.cime/config_batch.xml file to get a truncated version of the job_id into the job submission script. But I don't know how... Is anyone able to help?Thanks,Ingrid
 

inos@bas_ac_uk

Ingrid Cnossen
Member
Just in case anyone else is running into this problem and is struggling to find a way around it, I found a solution. In my ~/.cime/config_batch.xml file I specified an extra -N option (job name option on PBS) which overrrides the original -N option that is automatically generated. I added the following directive: -N $ENV{job_id_short}In this case job_id_short is an environment variable I specified, but it could be anything, as long as it is no more than 15 characters long...Ingrid  
 
Top