search

ESMCI / cime

Common Infrastructure for Modeling the Earth
http://esmci.github.io/cime
Other
161 stars 206 forks source link

Outline and practice steps for bringing new CIME to ACME #228

Closed rljacob closed 7 years ago

rljacob commented 8 years ago

I've put together wiki page to start listing the steps needed to bring the new CIME in to ACME. https://github.com/ACME-Climate/cime/wiki/Converting-ACME-from-cime2-to-ESMCI-cime

This issue can be used for discussion and also for reporting on merge attempts.

rljacob commented 8 years ago

Attempt on 6/12/16. Use ESMCI f8417e7b795604bb0 and ACME 3c24d7c37c4f2.

clone latest ACME; make a branch move cime to cime.old: git mv cime cime.old commit above Add new cime: git subtree add --squash --prefix=cime/ f8417e7b795604bb0 Going to try an I-case first so finish creating of cime_config for clm. cd to components/clm/bld git mv clm.buildlib ../cime_config/buildlib git mv clm.buildnml ../cime_config/buildnml git mv user_nl_clm ../cime_config/

Try an I case. (will fail at river if all goes well with land since didn't finish cime_config for the river model). But got an error in create_newcase.

./create_newcase -res f19_f19 -case tryIcase -compset I1850CLM45CN -compiler intel
Compset longname is 1850_DATM%QIA_CLM45%CN_SICE_SOCN_RTM_SGLC_SWAV 
Compset specification file is /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/config_compsets.xml
Pes     specification file is /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/config_pes.xml
Traceback (most recent call last):
  File "./create_newcase", line 180, in <module>
    _main_func()
  File "./create_newcase", line 165, in _main_func
    walltime=walltime)
  File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../utils/python/CIME/case.py", line 447, in configure
    self._get_component_config_data()
  File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../utils/python/CIME/case.py", line 378, in _get_component_config_data
    compobj = Component(comp_config_file)
  File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../utils/python/CIME/XML/component.py", line 22, in __init__
    EntryID.__init__(self,infile)
  File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../utils/python/CIME/XML/entry_id.py", line 17, in __init__
    GenericXML.__init__(self, infile)
  File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../utils/python/CIME/XML/generic_xml.py", line 29, in __init__
    self.read(infile)
  File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../utils/python/CIME/XML/generic_xml.py", line 52, in read
    self.tree = ET.parse(infile)
  File "/soft/python/2.7.6/lib/python2.7/xml/etree/ElementTree.py", line 1182, in parse
    tree.parse(source, parser)
  File "/soft/python/2.7.6/lib/python2.7/xml/etree/ElementTree.py", line 656, in parse
    parser.feed(data)
  File "/soft/python/2.7.6/lib/python2.7/xml/etree/ElementTree.py", line 1642, in feed
    self._raiseerror(v)
  File "/soft/python/2.7.6/lib/python2.7/xml/etree/ElementTree.py", line 1506, in _raiseerror
    raise err
xml.etree.ElementTree.ParseError: mismatched tag: line 215, column 4
Exit 1
rljacob commented 8 years ago

@mvertens is there a problem with the config_compsets.xml for clm?

jedwards4b commented 8 years ago

Yes, minor changes will need to be made in all of the component models config_component.xml files

On Tue, Jul 12, 2016 at 5:58 PM, Robert Jacob notifications@github.com wrote:

@mvertens https://github.com/mvertens is there a problem with the config_compsets.xml for clm?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ESMCI/cime/issues/228#issuecomment-232216903, or mute the thread https://github.com/notifications/unsubscribe/AF16GGfafVaDZHTFgy4hgNqH3XKRqKTsks5qVCoegaJpZM4I6LR7 .

Jim Edwards

CESM Software Engineer National Center for Atmospheric Research Boulder, CO

mvertens commented 8 years ago

@rjacob i assume you are doing this with alm - not clm? i'm happy to help you sort out the problem with config_compsets.xml for clm - but i have no access to this code

On Tue, Jul 12, 2016 at 5:58 PM, Robert Jacob notifications@github.com wrote:

@mvertens https://github.com/mvertens is there a problem with the config_compsets.xml for clm?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ESMCI/cime/issues/228#issuecomment-232216903, or mute the thread https://github.com/notifications/unsubscribe/AHlxE1kvXbQrBuBpdR7MjJFBC5O0pkVOks5qVCofgaJpZM4I6LR7 .

rljacob commented 8 years ago

Yes its "alm" but its clm for all practical purposes. This is the config_compsets.xml from that tar file you made.

jedwards4b commented 8 years ago

Here is a list of differences in the clm cime_config xml files:


Index: config_component.xml
===================================================================
--- config_component.xml    (.../trunk_tags/clm4_5_9_r186/components/clm/cime_config)   (revision 79491)
+++ config_component.xml    (.../branches/clm4_5_8_r174_esmciport/components/clm/cime_config)   (revision 79491)
@@ -13,32 +13,6 @@
     <desc>Name of land component</desc>
   </entry>

-  <entry id="LND_PIO_REARRANGER">
-    <values>
-      <value compset="_CLM40" >2</value>
-      <value compset="_CLM45" >1</value>
-      <value compset="_CLM50" >1</value>
-    </values>
-  </entry>
-
-  <entry id="BARRIER_N">
-    <values>
-      <value compset="_DATM" >1</value>
-    </values>
-  </entry>
-
-  <entry id="BARRIER_OPTION">
-    <values>
-      <value compset="_DATM" >ndays</value>
-    </values>
-  </entry>
-
-  <entry id="ATM_PIO_TYPENAME">
-    <values>
-      <value compset="DATM.*_CLM" >netcdf</value>
-    </values>
-  </entry>
-
   <entry id="LND_TUNING_MODE">
     <type>char</type>
     <group>run_component_clm</group>
@@ -58,7 +32,7 @@
   <entry id="CLM_CONFIG_OPTS" >
     <type>char</type>
     <default_value></default_value>
-    <values additive='yes'>
+    <values modifier='additive'>
       <value compset="_CLM40"           >-phys clm4_0</value>
       <value compset="_CLM40%[^_]*CN"   >-bgc cn</value>
       <value compset="_CLM40%[^_]*CNDV" >-bgc cndv</value>
Index: config_compsets.xml
===================================================================
--- config_compsets.xml (.../trunk_tags/clm4_5_9_r186/components/clm/cime_config)   (revision 79491)
+++ config_compsets.xml (.../branches/clm4_5_8_r174_esmciport/components/clm/cime_config)   (revision 79491)
@@ -13,7 +13,7 @@
     TIME_ATM[%phys]_LND[%phys]_ICE[%phys]_OCN[%phys]_ROF[%phys]_GLC[%phys]_WAV[%phys][_BGC%phys]
     Where for the CAM specific compsets below the following is supported
     TIME = Time period (e.g. 2000, HIST, RCP8...)
-    ATM  = [CAM4, CAM5]
+    ATM  = [CAM40, CAM50, CAM55]
     LND  = [CLM40, CLM45, CLM50, SLND]
     ICE  = [CICE, DICE, SICE]
     OCN  = [DOCN, ,AQUAP, SOCN]
@@ -566,6 +566,18 @@
       </values>
     </entry>

+  <entry id="BARRIER_N">
+    <values>
+      <value compset="_DATM" >1</value>
+    </values>
+  </entry>
+
+  <entry id="BARRIER_OPTION">
+    <values>
+      <value compset="_DATM" >ndays</value>
+    </values>
+  </entry>
+
   </entries>

 </compsets>
rljacob commented 8 years ago

The problem seems to be in our new config_components.xml for alm. If I use the cesm1.5b6 version (and remove LND_PIO_REARRANGER) it gets past the create_newcase!

rljacob commented 8 years ago

Does the last line from the error message: xml.etree.ElementTree.ParseError: mismatched tag: line 215, column 4 mean that there's just a syntax error?

jedwards4b commented 8 years ago

Yes - you can check the validity of an xml file in a number of ways such as using xmllint or emacs nxml mode

rljacob commented 8 years ago

Looks like we have xmllint on blues. What arguments should I use?

rljacob commented 8 years ago

I put the config_components.xml on gist https://gist.github.com/rljacob/10b004ac59edea495e7158097142c81d

The output of xmllint without any args seems helpful but the tags its complaining about seem fine.

blogin2[180]: xmllint config_component.xml config_component.xml:215: parser error : expected '>' ^ config_component.xml:225: parser error : Opening and ending tag mismatch: description line 191 and definitions_variables ^ config_component.xml:226: parser error : Premature end of data in tag definitions_variables line 5 Exit 1 blogin2[181]

jedwards4b commented 8 years ago

line 206 is missing a closing </desc>

mvertens commented 8 years ago

The LND_PIO_REARRANGER should not have been in there. The BARRIERS should be moved to clm config_compsets.xml not in the clm config_components.xml. If you updated to the latest clm trunk tag - this is a problem and has been fixed in the svn esmci branch. See https://svn-ccsm-models.cgd.ucar.edu/clm2/branches/clm4_5_8_r174_esmciport/components/clm

On Wed, Jul 13, 2016 at 11:10 AM, Robert Jacob notifications@github.com wrote:

The problem seems to be in our new config_components.xml for alm. If I use the cesm1.5b6 version (and remove LND_PIO_REARRANGER) it gets past the create_newcase!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ESMCI/cime/issues/228#issuecomment-232422026, or mute the thread https://github.com/notifications/unsubscribe/AHlxE-b6WhFphwlVe4vq3iwdAhKJESLMks5qVRvqgaJpZM4I6LR7 .

rljacob commented 8 years ago

Thanks @jedwards4b! Is there a way xmllint or emacs could point me directly to line 206?

jedwards4b commented 8 years ago

emacs was pointing me to the parent closing and saying it was in error, i moved it up line by line until it wasn't in error anymore to find the real problem...

rljacob commented 8 years ago

With the fix to config_components.xml, create_newcase competes. Now for "case.setup" there's a different error:

preview_namelists Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../components/data_comps/datm/cime_config/buildnml Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml ERROR: Command: '/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase' failed with error 'Can't locate SetupTools.pm in @INC (@INC contains: /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase/Tools /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5 .) at /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml line 13.'

However @mvertens told me we need a new buildnml to work with CIME5 and supplied one.

rljacob commented 8 years ago

Using the clm buildnml suppled by @mvertens, I get a different error:

Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../components/data_comps/datm/cime_config/buildnml Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml ERROR: Command: '/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase' failed with error '*\ can't open file: /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase/user_nl_clm at /lcrc/group/earthscience/jacob/ACME2/cime/utils/perl5lib/Config/SetupTools.pm line 31' Exit 1 blogin3[123]:

Not sure why user_nl_clm was not copied from components/clm/cime_config.

I copied it by hand and got a different error: preview_namelists Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../components/data_comps/datm/cime_config/buildnml Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml ERROR: Command: '/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildnml /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase' failed with error 'ERROR : CLM build-namelist::CLMBuildNamelist::check_for_perl_utils() : Cannot find perl module "SetupTools.pm" in directory "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools"

ERROR clm.buildnml: /lcrc/group/earthscience/jacob/ACME2/components/clm/bld/build-namelist -infile /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase/Buildconf/clmconf/namelist -csmdata /home/ccsm-data/inputdata -inputdata /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase/Buildconf/clm.input_data_list -ignore_ic_year -namelist " &clm_inparm start_ymd=00010101 /" -use_case 1850_control -res 1.9x2.5 -clm_start_type default -envxml_dir /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase -l_ncpl 48 -lnd_frac /home/ccsm-data/inputdata/share/domains/domain.lnd.fv1.9x2.5_gx1v6.090206.nc -glc_nec 0 -co2_ppmv 284.7 -co2_type constant -config /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase/Buildconf/clmconf/config_cache.xml -bgc cn failed: 512' Exit 1

rljacob commented 8 years ago

I just had to change the location of SetupTools.pm in CLMBuildNamelist.pm. preview_namelists made it to rtm which I did not finish converting.

Finish it by moving current ACME versions as follows: git mv rtm.buildlib ../cime_config/buildlib git mv rtm.buildnml ../cime_config/buildnml git mv user_nl_rtm ../cime_config

Had to edit buildnml to change location of SetupTool.pm and change all use of CCSMROOT to SRCROOT.

case.setup works for the I case!

rljacob commented 8 years ago

Next is case.build. The buidlib failed for land: /lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lnd.bldlog.160713-180158

message: ar: /lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lnd/obj/lib/lib/lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lib.a: No such file or directory gmake: *** [/lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lnd/obj/lib/lib/lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lib.a] Error 1 ERROR: clm.buildlib gmake complib -j 4 MODEL=clm COMPLIB=/lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lnd/obj/lib/lib/lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lib.a USER_CPPDEFS=" " -f /lcrc/group/earthscience/jacob/ACME2/cime/scripts/tryIcase/Tools/Makefile failed: 512

Also all the land .o and .mod files appear in the $CASEROOT.

buildlib isn't different between ACME cime2 and cesm1.5b6 as far as I can tell.

rljacob commented 8 years ago

Any ideas why the land is building inside $CASEROOT? The buildlib I'm using is identical to clm4_5_9_r186

mvertens commented 8 years ago

Could you please try to use either the following buildlib -(now in python) or modify it to be usable for alm? I cleaned up the buildlib.py mechanism and its in interaction with clm in a previous PR. https://svn-ccsm-models.cgd.ucar.edu/clm2/branches/clm4_5_8_r174_esmciport/components/clm/cime_config/buildlib

On Thu, Jul 14, 2016 at 10:20 PM, Robert Jacob notifications@github.com wrote:

Any ideas why the land is building inside $CASEROOT? The buildlib I'm using is identical to clm4_5_9_r186

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ESMCI/cime/issues/228#issuecomment-232856530, or mute the thread https://github.com/notifications/unsubscribe-auth/AHlxE5vY2hUh8-0mzrd1VNcOV8k9tsblks5qVwqQgaJpZM4I6LR7 .

rljacob commented 8 years ago

That worked much better. Now the only error is at the very end: err=gmake: * No rule to make target /lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/intel/mvapich/nodebug/nothreads/MCT/noesmf/lib/libclm.a', needed by/lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/acme.exe'. Stop. gmake: * Waiting for unfinished jobs....

libclm.a exists but its in /lcrc/project/ACME/jacob/acme_scratch/tryIcase/bld/lib with the other model libs.

jedwards4b commented 8 years ago

What is in the variable CLM_CONFIG_OPTS in env_build.xml?

rljacob commented 8 years ago

`

char Provides option(s) for the CLM configure utility. CLM_CONFIG_OPTS are normally set as compset variables (e.g., -bgc cn) and in general should not be modified for supported compsets. It is recommended that if you want to modify this value for your experiment, you should use your own user-defined component sets via using create_newcase with a compset_file argument. This is an advanced flag and should only be used by expert users.

`

jedwards4b commented 8 years ago

The issues is in cime_config/acme/machines/Makefile starting at about line 695

rljacob commented 8 years ago

The equivalent logic in the cesm Makefile is: CLMVER = $(filter $(CLM_CONFIG_OPTS), clm5_0 clm4_5) ifeq ($(CLMVER),$(null)) LNDOBJDIR = $(EXEROOT)/lnd/obj LNDLIBDIR=$(LIBROOT) LNDLIB := liblnd.a else LNDOBJDIR = $(SHAREDPATH)/$(COMP_INTERFACE)/$(ESMFDIR)/clm/obj LNDLIBDIR = $(SHAREDPATH)/$(COMP_INTERFACE)/$(ESMFDIR)/lib LNDLIB := libclm.a endif INCLDIR += -I$(LNDOBJDIR)

This seems to say the library should be in $(SHAREDPATH)/$(COMP_INTERFACE)/$(ESMFDIR)/lib for clm4_5. Why is that?

Adding that logic doesn' help. The libclm.a is still built in $LIBROOT

jedwards4b commented 8 years ago

Because most everything in clm4_5 and clm5_0 is runtime configurable so we don't need to rebuild it for every test that uses it. Putting it in the shared library path allows multiple tests to use the same build. Not sure why you didn't want to do that in acme too.

On Fri, Jul 15, 2016 at 4:33 PM, Robert Jacob notifications@github.com wrote:

The equivalent logic in the cesm Makefile is: CLMVER = $(filter $(CLM_CONFIG_OPTS), clm5_0 clm4_5) ifeq ($(CLMVER),$(null)) LNDOBJDIR = $(EXEROOT)/lnd/obj LNDLIBDIR=$(LIBROOT) LNDLIB := liblnd.a else LNDOBJDIR = $(SHAREDPATH)/$(COMP_INTERFACE)/$(ESMFDIR)/clm/obj LNDLIBDIR = $(SHAREDPATH)/$(COMP_INTERFACE)/$(ESMFDIR)/lib LNDLIB := libclm.a endif INCLDIR += -I$(LNDOBJDIR)

This seems to say the library should be in $(SHAREDPATH)/$(COMP_INTERFACE)/$(ESMFDIR)/lib for clm4_5. Why is that?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ESMCI/cime/issues/228#issuecomment-233085068, or mute the thread https://github.com/notifications/unsubscribe-auth/AF16GP0CaEJt9_TsBboj8QRRxVDjOybqks5qWArIgaJpZM4I6LR7 .

Jim Edwards

CESM Software Engineer National Center for Atmospheric Research Boulder, CO

rljacob commented 8 years ago

I'm trying to put it in shared but it seems to be ignoring me. CLM_CONFIG_OPTS is clm4_5 so the LND*DIR should be the shared. It looks in shared at the end but still builds in $EXEROOT/lnd/obj and puts the library in $EXEROOT/lib. What am I missing?

jedwards4b commented 8 years ago

I think that the problem may be at line 508 of utils/python/CIME/build.py:

   if comp_lnd == "clm" and not "clm4_0" in clm_config_opts:

is your comp_lnd now alm instead of clm?

rljacob commented 8 years ago

That fixed the build! I changed valid_values and default_value from "alm" to "clm" in COMP_LND entry in config_component.xml.

rljacob commented 8 years ago

New issue: I tried running with case.submit and got an error message: "ERROR: must set CIMEROOT environment variable" CIMEROOT is defined correctly in env_case.xml.

I submitted the case.run script directly using qsub and got the same error.

jedwards4b commented 8 years ago

you need to set CIMEROOT in the environment on the compute node, on most systems this just means setting it on the front end node, but some machines require an argument to the submit command - what machine and queueing system is it?

rljacob commented 8 years ago

Blues and pbs. Something must have broken in the blues config. I get the same error with an X case.

rljacob commented 8 years ago

After upgrading to latest ESMCI cime master (92a5d03244f096038da32b202637cfab105336b0) and starting over, got a new error in the clm buildlib (which is a copy of the one from @mvertens branch above):

Traceback (most recent call last): File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildlib", line 88, in _main_func() File "/lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/clm/cime_config/buildlib", line 79, in _main_func rc, out, err = run_cmd(cmd, ok_to_fail=True) TypeError: run_cmd() got an unexpected keyword argument 'ok_to_fail'

jgfouca commented 8 years ago

@rljacob there was a minor API change to run_cmd. Any run_cmd that had ok_to_fail needs to remove the ok_to_fail arg. All other run_cmd calls need to be changed to call run_cmd_no_fail

rljacob commented 8 years ago

That worked! And I can now run an I case on blues with the latest master. Next: an IM case (with mosart).

rljacob commented 8 years ago

IM case works. F case is next.

rljacob commented 8 years ago

For an F-case, I'm doing ./create_newcase -case tryF19 -compset FC5 -res f19_f19 -compiler intel I've renamed/moved the ACME cam.buildlib and cam.builnml and made the necessary perl path and variable name changes.

create_newcase works but case.setup gives this error:
Running /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/cam/cime_config/buildnml
ERROR: Command /lcrc/group/earthscience/jacob/ACME2/cime/scripts/Tools/../../../components/cam/cime_config/buildnml failed rc=2
out=ERROR: unrecognized arguments: 1.9x2.5
err=SYNOPSIS
     configure [options]

....(print out of the configure help page deleted).....

ERROR cam.buildnml: /lcrc/group/earthscience/jacob/ACME2/components/cam/bld/configure -s -ccsm_seq -ice none -ocn docn -comp_intf mct  -spmd -spmd -smp -nosmp -dyn fv -dyn_target  -res 1.9x2.5   -phys cam5 failed: 65280

Exit 1 blogin2[182]:

jedwards4b commented 8 years ago

It seems that the cams configure options are not consistent with what buildnml expects them to be.

gold2718 commented 8 years ago

Where is -dyn_target getting set? There is supposed to be a string after it (preqx or preqx_acc).

jedwards4b commented 8 years ago

Ahh dyn_target is not used in cesm cam - i think you'll need to modify your buildnml to handle it

rljacob commented 8 years ago

But I'm using the ACME cam buildnml which should work with the ACME cam configure. I just moved "cam.buildnml" to the right place and changed the name, perl paths, etc.

gold2718 commented 8 years ago

If you look in buildnml, it is using the value of $CAM_TARGET which appears to be blank (unset). Who is supposed to set that in ACME?

rljacob commented 8 years ago

Ah that was set in create_newcase and we took it out for cime5:

if ($camse_target eq "preqx") { $cfg_ref->set('CAM_TARGET', "preqx"); } elsif ($camse_target eq "preqx_acc") { $cfg_ref->set('CAM_TARGET', "preqx_acc"); } else { die "ERROR: option -camse_target must be preqx (default) or preqx_acc" }

The default value was in config_definitions and didn't get moved to our new cam config_components.

rljacob commented 8 years ago

I got past the CAM part of preview_namelists by hard-coding CAM_TARGET in buildnml. Now working on cice which didn't get its config_components.xml created.

rljacob commented 8 years ago

F, G and D cases should all work. mpas-o and mpas-cice have been added. Will now try a fully coupled case.

rljacob commented 8 years ago

Trying to do an IG case: clm with MPAS-landice. Getting this error from CLM-Buildnamelist:

err=ERROR : CLM build-namelist::CLMBuildNamelist::setup_logic_glacier() : glc_smb set to .true. does NOT agree with -glc_smb argument of .. (set with GLC_SMB env variable)

Not sure how the argument is getting set to ".."

rljacob commented 8 years ago

Nevermind. Found the problem. GLC_SMB was missing from config_component.xml in mpasli. Then had to re-create the case from scratch.

rljacob commented 7 years ago

Final steps:

  1. make sure the cime_config directories in ACME master have content euqivalent to latest versions from the uptocime5.2 branch.
  2. Try reverting the coupler and share code on uptocime5.2 to the current ACME versions so get around BFB issues.
  3. Bring recent developments from ACME/cime in to the ESMCI maint-cime5.1 branch.
  4. Make a new branch off ACME master and bring in the head of the main-cime5.1 branch
    1. update ACME/cime
    2. go through each components bld directory and a. remove the buildlib and .buidnml and .cpl7.template files. b. Move bld/usernl to cime_config (if not done already) c. copy from the uptocime5.2 branch any needed mods for build-namelist, configure
    3. Run testing.
rljacob commented 7 years ago

The following cime_config files on the branch are more up-to-date compared to what is on master: cice/cime_config/config_component.xml clm/cime_config/buildnml mpasli/cime_config/config_component.xml