search

E3SM-Project / E3SM

Energy Exascale Earth System Model source code. NOTE: use "maint" branches for your work. Head of master is not validated.
https://docs.e3sm.org/E3SM
Other
343 stars 352 forks source link

maint-1.2 not supported on Compy #3904

Open worleyph opened 3 years ago

worleyph commented 3 years ago

@rljacob indicated that all maint tags should work on all supported systems. On compy using maint-1.2,

 ./create_newcase --case FC5AV1C-L_ne4_ne4_m1.2 --res ne4_ne4 --compset FC5AV1C-L --machine compy --compiler intel --project e3sm
 Compset longname is 2000_CAM5%AV1C-L_CLM45%SPBC_CICE%PRES_DOCN%DOM_SROF_SGLC_SWAV
 Compset specification file is /qfs/people/worl312/E3SM/MAINT/maint-1.2/E3SM/cime/../components/cam/cime_config/config_compsets.xml
 Compset forcing is 
 Com forcing is present day:
 ATM component is cam5 physics:CAM with complete set of E3SM atmospheric mods for V1 (72 layers model):CAM with complete set of E3SM atmospheric mods for V1 (72 layers model) and ACES4BGC SOAG emissions- v04p2:
 LND component is clm4.5 physics:clm4.5 Satellite phenology:clm4.5 Satellite phenology with black carbon deposition:
 ICE component is prescribed cice:
 OCN component is DOCN   prescribed ocean mode
 ROF component is Stub river component
 GLC component is Stub glacier (land ice) component
 WAV component is Stub wave component
 ESP component is 
 Pes     specification file is /qfs/people/worl312/E3SM/MAINT/maint-1.2/E3SM/cime/config/e3sm/allactive/config_pesall.xml
 ERROR: No machine compy found

In contrast,

worleyph commented 3 years ago

Assigned to @rljacob for reassignment to appropriate personnel, if it is in fact important that maint-1.2 run on compy.

bibiraju commented 3 years ago

@rljacob @worleyph I am able to successfully run a case (FC5AVIC-L and ne4_ne4) using maint-1.2 branch on Compy with the following changes.

  1. Edited config_pesall.xml to add pes config for Compy as in maint-1.0
  2. Edited config_batch.xml to add compy slurm queues as in maint-1.0
  3. Edited config_compilers.xml to add Compy compilers as in maint-1.0
  4. Edited config_machines.xml to add Compy as in maint-1.0
  5. Edited provenance.py to include compy as in maint-1.0
  6. Added syslog.compy file as in maint-1.0

I will do more testing with other cases as well. Is this what we are looking for?

worleyph commented 3 years ago

@bibiraju , sounds right to me. I'd like to try your update myself, just to check that the performance archiving logic is working correctly. Do you have a branch that I could check out, or can you point me at your sandbox?

@rljacob , is there a test suite for the maint branches that @bibiraju can use to verify correctness?

bibiraju commented 3 years ago

@worleyph here's my branch bibiraju/maint-1.2/compy https://github.com/E3SM-Project/E3SM/tree/bibiraju/maint-1.2/compy

rljacob commented 3 years ago

e3sm-developer should work on maint-1.2.

@bibiraju yes you are on the right track. You just need to backport the compy settings from master to maint-1.2.

worleyph commented 3 years ago

@bibiraju , repeated the same F case and also worked for me. Looked closely at performance archiving logic, including checkpointing, and it is all working fine. Thanks!

bibiraju commented 3 years ago

e3sm-developer should work on maint-1.2.

@bibiraju yes you are on the right track. You just need to backport the compy settings from master to maint-1.2.

@rljacob should I match the master branch or maint-1.0?

rljacob commented 3 years ago

master.

worleyph commented 3 years ago

@rljacob , does that mean that maint-1.0 and maint-1.1 should (eventually) be resynced to master? A lot has changed in the CIME layer since then.

rljacob commented 3 years ago

No we never actually merge master to the maint branches. But the latest settings for compy from master should be on the maint branches (ported by hand).

worleyph commented 3 years ago

@bibiraju and @rljacob , can this be closed?