fates-interface: integration of external model #1536

rgknox · 2017-05-14T23:26:16Z

This change set completes a first version of 1) the ALM-side interface for FATES,
2) initializes FATES as a submodule and 3) adds a test to acme_land_developer.
Restart simulations, while having the code in place, are not passing tests yet and
will be addressed in forthcoming changesets. Modifications to the namelist build
system were necessary. Thus, there are trivial conflicts with namelists generated
by master, specifically use_ed=.false. is now appended to all namelists where it
is not used.This has no impact on existing tests' model output, however CLM
regression tests will fail namelist comparisons for this reason.
As this PR initializes a new submodule, all CLM users must update their submodule
pointers to bring the fates codeset in, without this the code will not compile.

[BFB]
[NML]

rgknox · 2017-05-14T23:26:37Z

This PR replaces #1531.

rljacob · 2017-05-15T04:01:08Z

Does this depend on #1460

bishtgautam · 2017-05-15T14:59:19Z

@rljacob: This PR doesn't depend on #1460

bishtgautam

Few minor changes

bishtgautam · 2017-05-15T15:17:06Z

components/clm/bld/CLMBuildNamelist.pm

-       # ED is not a clm4_0 option and should not be used with crop and not with clm4_0
-       fatal_error("** Cannot turn betr mode on with crop or with clm4_0 physics.\n" );
-    }
+#    if ( $nl_flags->{'ed_mode'} == 1 ) {


The original logic in the if-loop was incorrect and should be fixed by

if ( $nl_flags->{$var} == 1 ) { # BeTR is not a clm4_0 option and should not be used with crop and not with clm4_0 fatal_error("** Cannot turn betr mode on with crop or with clm4_0 physics.\n" ); }

bishtgautam · 2017-05-15T17:46:57Z

components/clm/src/main/clm_instMod.F90

@@ -111,13 +111,11 @@ module clm_instMod
  type(phosphorusflux_type)                           :: phosphorusflux_vars
  type(clm_bgc_interface_data_type)                   :: clm_bgc_data
  type(chemstate_type)                                :: chemstate_vars
-! (FATES-INTERF)
-!  type(hlm_fates_interface_type)                      :: clm_fates (FATES-INTERF)
+  type(hlm_fates_interface_type)                      :: clm_fates


Rename clm_fates to alm_fates.

bishtgautam · 2017-05-15T17:52:50Z

components/clm/src/main/ncdio_pio.F90.in

@@ -655,8 +661,6 @@ contains

  end subroutine ncd_inqvdlen_byDesc

-
-  !-----------------------------------------------------------------------
  subroutine ncd_inqvdlen_byName(ncid,varname,dimnum,dlen,err_code)


Add back the two deleted lines:

!-----------------------------------------------------------------------

will do, I think those got deleted in the rebasing effort, there were some conflicts in there if I remember correctly, and the way the head vs base diffs were structured was a little confusing.

bishtgautam · 2017-05-17T17:10:02Z

@rgknox: Following changes are needed to the first commit message:

Add [BFB] and [NML] at the end of the message, and
Break the commit message into multiple lines instead of a single
long line because that message is copied & pasted in the merge
commit.

jqyin

Looks good. The threading is in initialization part and not nested threading in the main driver calls.

…t (PR #1536) This change set completes a first version of 1) the ALM-side interface for FATES, 2) initializes FATES as a submodule and 3) adds a test to acme_land_developer. Restart simulations, while having the code in place, are not passing tests yet and will be addressed in forthcoming changesets. Modifications to the namelist build system were necessary. Thus, there are trivial conflicts with namelists generated by master, specifically use_ed=.false. is now appended to all namelists where it is not used.This has no impact on existing tests' model output, however CLM regression tests will fail namelist comparisons for this reason. As this PR initializes a new submodule, all CLM users must update their submodule pointers to bring the fates codeset in, without this the code will not compile. [BFB] [NML]

bishtgautam · 2017-05-19T16:48:13Z

@amametjanov reported following error on Cetus for SMS_Lm1.f45_f45.ICLM45ED.cetus_ibm.clm-fates

4: "/gpfs/mira-home/azamatm/repos/ACME-test/components/clm/src/external_models/fates/biogeochem/EDCohortDynamicsMod.F90", line 837: 1525-109 Error encountered while attempting to deallocate a data object.  The program will stop.
2017-05-19 16:30:32.081 (WARN ) [0xfff8264ca70] CET-00000-11331-128:2129189:ibm.runjob.client.Job: normal termination with status 1 from rank 4```

@rgknox: Do you have access to Cetus to debug this failure?

This change set completes a first version of 1) the ALM-side interface for FATES, 2) initializes FATES as a submodule and 3) adds a test to acme_land_developer. Restart simulations, while having the code in place, are not passing tests yet and will be addressed in forthcoming changesets. Modifications to the namelist build system were necessary. Thus, there are trivial conflicts with namelists generated by master, specifically use_ed=.false. is now appended to all namelists where it is not used.This has no impact on existing tests' model output, however CLM regression tests will fail namelist comparisons for this reason. As this PR initializes a new submodule, all CLM users must update their submodule pointers to bring the fates codeset in, without this the code will not compile. [BFB] [NML]

bishtgautam · 2017-05-19T21:28:40Z

The duration of the ICLM45ED test has been decreased to avoid failure on Cetus.
@rgknox has requested an account for ALCC to debug this failure in coming days.
This updated PR has been re-merged to next.

)" This reverts commit af9d2dd, reversing changes made to 216511d.

…into next (PR #1536)" This reverts commit 16bb80b, reversing changes made to db6993d.

This change set completes a first version of 1) the ALM-side interface for FATES, 2) initializes FATES as a submodule and 3) adds a test to acme_land_developer. Restart simulations, while having the code in place, are not passing tests yet and will be addressed in forthcoming changesets. Modifications to the namelist build system were necessary. Thus, there are trivial conflicts with namelists generated by master, specifically use_ed=.false. is now appended to all namelists where it is not used.This has no impact on existing tests' model output, however CLM regression tests will fail namelist comparisons for this reason. As this PR initializes a new submodule, all CLM users must update their submodule pointers to bring the fates codeset in, without this the code will not compile. [BFB] [NML]

rljacob

Should have reviewed this sooner. All the changes to files in cime need to be in a separate commit. This is because cime is an external.

This reverts commit 9233c4b, reversing changes made to 1793e10.

rgknox · 2017-05-24T00:50:37Z

ok, I should be able to do an interactive rebase, and then go into edit mode to split this up.

Or, I guess I can just git reset HEAD~ and go from there.

rgknox · 2017-05-24T03:24:39Z

Commits have been split into 3 parts. The second two are cime side, the first is components ie non-cime.

For comparison I copied and renamed the single commit branch here:

https://github.com/rgknox/ACME/tree/rgknox/lnd/fates-part3-pre-split

rljacob · 2017-05-24T03:44:06Z

Looks good. Gautum, you can re-merge this to next.

This change set completes a first version of 1) the ALM-side interface for FATES, 2) initializes FATES as a submodule and 3) adds a test to acme_land_developer. Restart simulations, while having the code in place, are not passing tests yet and will be addressed in forthcoming changesets. Modifications to the namelist build system were necessary. Thus, there are trivial conflicts with namelists generated by master, specifically use_ed=.false. is now appended to all namelists where it is not used.This has no impact on existing tests' model output, however CLM regression tests will fail namelist comparisons for this reason. As this PR initializes a new submodule, all CLM users must update their submodule pointers to bring the fates codeset in, without this the code will not compile. [BFB] [NML]

rgknox added BFB PR leaves answers BFB Land labels May 14, 2017

rgknox assigned bishtgautam May 14, 2017

rgknox requested review from rljacob, bishtgautam and jqyin May 14, 2017 23:26

bishtgautam reviewed May 16, 2017

View reviewed changes

bishtgautam approved these changes May 17, 2017

View reviewed changes

jqyin approved these changes May 17, 2017

View reviewed changes

rgknox force-pushed the rgknox/lnd/fates-part3 branch 2 times, most recently from 82d5c18 to 1ee1d0a Compare May 23, 2017 20:50

bishtgautam pushed a commit that referenced this pull request May 23, 2017

Revert "Merge branch 'rgknox:rgknox/lnd/fates-part3' into next (PR #1536

2d5aa5b

)" This reverts commit af9d2dd, reversing changes made to 216511d.

bishtgautam pushed a commit that referenced this pull request May 23, 2017

Revert "Merge remote-tracking branch 'rgknox/rgknox/lnd/fates-part3' …

1793e10

…into next (PR #1536)" This reverts commit 16bb80b, reversing changes made to db6993d.

bishtgautam mentioned this pull request May 23, 2017

Updates VSFM and adds EMI #1460

Merged

rljacob requested changes May 23, 2017

View reviewed changes

bishtgautam pushed a commit that referenced this pull request May 24, 2017

Revert "Merge branch 'rgknox/lnd/fates-part3' into next (PR #1536)"

e79eb18

This reverts commit 9233c4b, reversing changes made to 1793e10.

rgknox added 3 commits May 23, 2017 19:57

fates-interface: added fates submodule, w/o cime components

08b728d

fates-interface: added machine configuration and tests to cime

a8fdda3

fates-interface: testmod for fates

86c72aa

rgknox force-pushed the rgknox/lnd/fates-part3 branch from 1ee1d0a to 86c72aa Compare May 24, 2017 03:18

rljacob approved these changes May 24, 2017

View reviewed changes

bishtgautam merged commit 86c72aa into E3SM-Project:master May 26, 2017

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fates-interface: integration of external model #1536

fates-interface: integration of external model #1536

rgknox commented May 14, 2017 •

edited by bishtgautam

Loading

rgknox commented May 14, 2017

rljacob commented May 15, 2017

bishtgautam commented May 15, 2017

bishtgautam left a comment

bishtgautam May 15, 2017

bishtgautam May 15, 2017

bishtgautam May 15, 2017

rgknox May 17, 2017

bishtgautam commented May 17, 2017

jqyin left a comment

bishtgautam commented May 19, 2017

bishtgautam commented May 19, 2017

rljacob left a comment

rgknox commented May 24, 2017 •

edited

Loading

rgknox commented May 24, 2017

rljacob commented May 24, 2017

fates-interface: integration of external model #1536

fates-interface: integration of external model #1536

Conversation

rgknox commented May 14, 2017 • edited by bishtgautam Loading

rgknox commented May 14, 2017

rljacob commented May 15, 2017

bishtgautam commented May 15, 2017

bishtgautam left a comment

Choose a reason for hiding this comment

bishtgautam May 15, 2017

Choose a reason for hiding this comment

bishtgautam May 15, 2017

Choose a reason for hiding this comment

bishtgautam May 15, 2017

Choose a reason for hiding this comment

rgknox May 17, 2017

Choose a reason for hiding this comment

bishtgautam commented May 17, 2017

jqyin left a comment

Choose a reason for hiding this comment

bishtgautam commented May 19, 2017

bishtgautam commented May 19, 2017

rljacob left a comment

Choose a reason for hiding this comment

rgknox commented May 24, 2017 • edited Loading

rgknox commented May 24, 2017

rljacob commented May 24, 2017

rgknox commented May 14, 2017 •

edited by bishtgautam

Loading

rgknox commented May 24, 2017 •

edited

Loading