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

Configuring CESM2 in CRAY XT50

nicolas cruz

nicase03
New Member
Hi, I'm configuring a new case for CESM 2.2 but always crash in this point.
ERROR: Command: '/usr/bin/xmllint --xinclude --noout --schema /scratchin/grupos/bam/home/nicolas.cruz/CESM-master/my_cesm_sandbox/cime/config/xml_schemas/config_workflow.xsd /scratchin/grupos/bam/home/nicolas.cruz/CESM-master/my_cesm_sandbox/cime/config/cesm/machines/config_workflow.xml' failed with error '/scratchin/grupos/bam/home/nicolas.cruz/CESM-master/my_cesm_sandbox/cime/config/cesm/machines/config_workflow.xml:55: element template: Schemas validity error : Element 'template': '$ENV{POSTPROCESS_PATH}/timeseries/template.timeseries' is not a valid value of the atomic type 'xs:anyURI'.
/scratchin/grupos/bam/home/nicolas.cruz/CESM-master/my_cesm_sandbox/cime/config/cesm/machines/config_workflow.xml:72: element template: Schemas validity error : Element 'template': '$ENV{POSTPROCESS_PATH}/timeseries/template.timeseries_transfer' is not a valid value of the atomic type 'xs:anyURI'

Any help thanks.
 

Attachments

  • new_case_report.txt
    3.8 KB · Views: 4
  • config_machines.txt
    4.3 KB · Views: 2

erik

Erik Kluzek
CSEG and Liaisons
Staff member
It's failing the XML schema validation for your config_workflow.xml file. It expects a field to be of a URI (Uniform Resource Identifier) type, which largely means it should be a webpage. So look at line 55 in that file and see what's going wrong.

Here's a web-page I found talking about the xs:anyURI datatype.

 

nicolas cruz

nicase03
New Member
Well, the problem was solved by modifying the config_workflow.xsd file by swapping anyURI for string:

job is case.run USER_REQUESTED_WALLTIME None USER_REQUESTED_QUEUE None WALLTIME_FORMAT %H:%M:%S
job is case.st_archive USER_REQUESTED_WALLTIME None USER_REQUESTED_QUEUE None WALLTIME_FORMAT %H:%M:%S
Creating Case directory /scratchin/grupos/bam/home/nicolas.cruz/CESM-master/my_cesm_sandbox/cime/scripts/nicolas.cruz/cases/b.e20.B1850.f19_g16.test

I don't know why USER_REQUESTED_WALLTIME None USER_REQUESTED_QUEUE None WALLTIME_FORMAT %H:%M:%S, anyway my newcase was created. Thanks
 

nicolas cruz

nicase03
New Member
It's failing the XML schema validation for your config_workflow.xml file. It expects a field to be of a URI (Uniform Resource Identifier) type, which largely means it should be a webpage. So look at line 55 in that file and see what's going wrong.

Here's a web-page I found talking about the xs:anyURI datatype.

Hi Erik
I Run the example of CESM Quick, b.e20.B1850.f19_g17.test, the model start but stop with this error.

POP aborting...
Stopping in ecosys_driver:ecosys_driver_init

How can i disable ecosystem in the model?
 

nicolas cruz

nicase03
New Member
Hi Erik
I Run the example of CESM Quick, b.e20.B1850.f19_g17.test, the model start but stop with this error.

POP aborting...
Stopping in ecosys_driver:ecosys_driver_init

How can i disable ecosystem in the model?
the problem is in POP
 

Attachments

  • ocn.log.62660.sdb.220523-125836.txt
    1.4 KB · Views: 6

mlevy

Michael Levy
CSEG and Liaisons
Staff member
The easiest way to turn MARBL off would be to remove ecosys from the OCN_TRACER_MODULES variable in env_build.xml. You can either run

Code:
$ ./xmlquery OCN_TRACER_MODULES

and then run ./xmlchange OCN_TRACER_MODULES="..." where the ... is everything from the xmlquery output except ecosys or just manually edit the file to remove ecosys. Note that you will have to rebuild the case after changing this variable:

Code:
$ ./case.build --clean-all
$ ./case.build

Note that MARBL is very expensive, so the default PE layout for your run might give the ocean more tasks than it needs (it keys off the presence of POP2%ECO in the compset long name), but it looks like you are running on your own cluster so that may not be an issue
 

nicolas cruz

nicase03
New Member
The easiest way to turn MARBL off would be to remove ecosys from the OCN_TRACER_MODULES variable in env_build.xml. You can either run

Code:
$ ./xmlquery OCN_TRACER_MODULES

and then run ./xmlchange OCN_TRACER_MODULES="..." where the ... is everything from the xmlquery output except ecosys or just manually edit the file to remove ecosys. Note that you will have to rebuild the case after changing this variable:

Code:
$ ./case.build --clean-all
$ ./case.build

Note that MARBL is very expensive, so the default PE layout for your run might give the ocean more tasks than it needs (it keys off the presence of POP2%ECO in the compset long name), but it looks like you are running on your own cluster so that may not be an issueThank
Thanks Michael, the model runs, but could not finish, everything indicates that it needs more time:

- Prestage required restarts into /lus/snx11029/e63116/CESM_master/project/e63116/BHIST.f09_g17.CMIP6-hist-noLu.003.test300/run
- Case input data directory (DIN_LOC_ROOT) is /lus/snx11029/e63116/CESM_master/inputdata
- Checking for required input datasets in DIN_LOC_ROOT
-------------------------------------------------------------------------
run command is aprun -n 336 -N 56 -d 1 /lus/snx11029/e63116/CESM_master/project/e63116/BHIST.f09_g17.CMIP6-hist-noLu.003.test300/bld/cesm.exe >> cesm.log.$LID 2>&1
=>> PBS: job killed: walltime 43305 exceeded limit 43200
 

nicolas cruz

nicase03
New Member
The easiest way to turn MARBL off would be to remove ecosys from the OCN_TRACER_MODULES variable in env_build.xml. You can either run

Code:
$ ./xmlquery OCN_TRACER_MODULES

and then run ./xmlchange OCN_TRACER_MODULES="..." where the ... is everything from the xmlquery output except ecosys or just manually edit the file to remove ecosys. Note that you will have to rebuild the case after changing this variable:

Code:
$ ./case.build --clean-all
$ ./case.build

Note that MARBL is very expensive, so the default PE layout for your run might give the ocean more tasks than it needs (it keys off the presence of POP2%ECO in the compset long name), but it looks like you are running on your own cluster so that may not be an issue
RESTART
why the model don`t create the rpointer.drv:
ERROR: CONTINUE_RUN is true but this case does not appear to have restart files staged in /lus/snx11029/e63116/CESM_master/project/e63116/b.e20.B1850.f19_g17.test102/run rpointer.drv
 
Top