Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

B1850 compset can no longer be created off of cime master #1532

Closed
mvertens opened this issue May 14, 2017 · 8 comments
Closed

B1850 compset can no longer be created off of cime master #1532

mvertens opened this issue May 14, 2017 · 8 comments
Assignees
Labels
Milestone

Comments

@mvertens
Copy link
Contributor

The PR with hash lbbbfc2
Merge pull request #1517 from jedwards4b/set_value_valid
is causing B1850 compsets to not work out of the box

The output is as follows:
Compset longname is 1850_CAM60_CLM50%BGC_CICE_POP2%ECO_MOSART_CISM2%NOEVOLVE_WW3_BGC%BDRD
Compset specification file is /glade/p/work/mvertens/cesm2_0_alpha06j.datamodels/cime/../cime_config/config_compsets.xml
Pes specification file is None
ERROR: No variable XROF_FLOOD_MODE found in case

This needs to be fixed as soon as possible since all PRs since then can no longer be used in the next cesm tag.

@billsacks billsacks added this to the cesm2 milestone May 14, 2017
@billsacks
Copy link
Member

I feel some responsibility for this, because I'm the one who pushed for getting this error-check put in place. Given the time crunch, I'd be reluctantly okay with backing out #1517 for now so that this error-check isn't done - as long as we plan to put it back in again soon when the cesm2 rush is done.

Although if we think that this is likely to be the only - or one of the only - offending xml variables, then it could be just as easy to fix this problem, I guess. My point is just that I'm not completely tied to the need to have the fix in #1517 right now if it's causing more trouble than it's worth.

@mvertens
Copy link
Contributor Author

mvertens commented May 14, 2017 via email

@jedwards4b
Copy link
Contributor

It's good to find and fix these now. Should there be an XROF_FLOOD_MODE variable? Probably not? I would rather fix this than back out the PR, I don't think it will take much. First thing Monday.

@ekluzek
Copy link
Contributor

ekluzek commented May 14, 2017

@jedwards4b I don't think that XROF_FLOOD_MODE should exist. FLOOD_MODE is an experimental configuration for CLM and an active ROF component (currently only RTM, but eventually MOSART). It routes an additional field through the coupler. I suppose having XROF_FLOOD_MODE could help to test this configuration. But, since it's a sideline mode right now I don't think you need it in the dead components. It should be easier to delete this rather than add it in and make sure it's working right.

@mvertens
Copy link
Contributor Author

mvertens commented May 15, 2017 via email

@jedwards4b
Copy link
Contributor

XROF_FLOOD_MODE is set in the config_grids.xml file in a line that looks like:

    <gridmap lnd_grid="1.9x2.5" rof_grid="r05" ocn_grid="gx1v7" >             
      <map name="XROF_FLOOD_MODE">ACTIVE</map>                                
    </gridmap>                                                                

I see two issues with this - first is this really where we should set this variable or should it be in the xrof config_component.xml file? Second - there are three grid attributes defined here, but if you look at the logic in grids.py it's only doing binary compares so if any two grids match we get a match. So in this case the lnd grid doesn't match but since rof and ocn do the variable is defined.

@jedwards4b
Copy link
Contributor

I don't understand why, if this variable is used with clm and rtm, it's defined in xrof?

@jedwards4b
Copy link
Contributor

I see these are also defined in xrof config_component.xml - I think that they should just be removed from the config_grids.xml file. I've also added a check in grids.py to make sure that the gridmap field in config_grids.xml has exactly two attributes. PR on the way.

jedwards4b added a commit that referenced this issue May 15, 2017
remove xrof_flood_mode from config_grids assure exactly two grids 
XROF_FLOOD_MODE as set in config_grids.xml was incorrectly being set for ROF modes which were not XROF, it also had three not two grid attributes.

Test suite: ERS.f09_g17.B1850.cheyenne_intel, scripts_regression_tests.py
Test baseline:
Test namelist changes:
Test status: bit for bit

Fixes #1532

User interface changes?:

Code review:mvertens
@ghost ghost removed the in progress label May 15, 2017
jgfouca pushed a commit that referenced this issue Oct 17, 2017
Fix bug in memory usage report and make work for Mira

Swap resident size and highwater size at the end of a run.  Wrong for all machines.
Print more detailed mem-usage to acme.log at the end of each model-day

Fix `GPTLget_memusage` on Mira/BGQ
Run CAM's `configure` with `-target_os bgq`, instead of defaulting to `linux` which sets `HAVE_SLASHPROC` macro
Num nodes given to cobalt should be `ceiling($TOTALPES/$PES_PER_NODE)`

[BFB]

* azamat/mira/fix-memusage:
  Switch memory highwater and usage in fortran writes
  Add CAM configure option target_os on BGQ
  Update GPTLget_memusage for Mira
jgfouca pushed a commit that referenced this issue Feb 23, 2018
Fix bug in memory usage report and make work for Mira

Swap resident size and highwater size at the end of a run.  Wrong for all machines.
Print more detailed mem-usage to acme.log at the end of each model-day

Fix `GPTLget_memusage` on Mira/BGQ
Run CAM's `configure` with `-target_os bgq`, instead of defaulting to `linux` which sets `HAVE_SLASHPROC` macro
Num nodes given to cobalt should be `ceiling($TOTALPES/$PES_PER_NODE)`

[BFB]

* azamat/mira/fix-memusage:
  Switch memory highwater and usage in fortran writes
  Add CAM configure option target_os on BGQ
  Update GPTLget_memusage for Mira
jgfouca pushed a commit that referenced this issue Mar 13, 2018
Fix bug in memory usage report and make work for Mira

Swap resident size and highwater size at the end of a run.  Wrong for all machines.
Print more detailed mem-usage to acme.log at the end of each model-day

Fix `GPTLget_memusage` on Mira/BGQ
Run CAM's `configure` with `-target_os bgq`, instead of defaulting to `linux` which sets `HAVE_SLASHPROC` macro
Num nodes given to cobalt should be `ceiling($TOTALPES/$PES_PER_NODE)`

[BFB]

* azamat/mira/fix-memusage:
  Switch memory highwater and usage in fortran writes
  Add CAM configure option target_os on BGQ
  Update GPTLget_memusage for Mira
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

4 participants