This adds additional write restart at end of run functionality. by dabail10 · Pull Request #984 · CICE-Consortium/CICE

dabail10 · 2024-10-17T20:42:37Z

For detailed information about submitting Pull Requests (PRs) to the CICE-Consortium,
please refer to: https://github.com/CICE-Consortium/About-Us/wiki/Resource-Index#information-for-developers

PR checklist

io_pio2/ice_restart.F90 has been modified to handle the multiple instances
nuopc/cmeps driver is modified

anton-seaice · 2024-10-17T21:00:52Z

There was already a "write_restart_at_endofrun" option added to the nuopc driver : - https://github.com/CICE-Consortium/CICE/pull/969/files

Is this doing the same thing ?

jedwards4b · 2024-10-17T21:05:23Z

Just making it functional with cesm.

jedwards4b · 2024-10-18T13:10:45Z


       !----------------
-       ! Restart alarm
+       ! Stop alarm


No - the reason they are reordered is so that a REST_OPTION=end can be initialized using the stop alarm. If you do the rest alarm first you can not initialize this option.

anton-seaice

This builds and runs

anton-seaice · 2024-10-20T23:37:42Z

         if (my_task == master_task) then
+#ifdef CESMCOUPLED
+            write(pointer_file,'(a,i4.4,a,i2.2,a,i2.2,a,i5.5)') &
+                 'rpointer.ice'//trim(inst_suffix)//'.',myear,'-',mmonth,'-',mday,'-',msec


Does this duplicate https://github.com/ESCOMP/CICE/blob/6d8c7f68c2ef80da9973f27c0077e54018bdbbfe/cicecore/cicedyn/general/ice_init.F90#L1202
?
which says:

#ifdef CESMCOUPLED pointer_file = trim(pointer_file) // trim(inst_suffix) #endif

No. That adds the inst_suffix which allows for single executable ensembles. This adds a date stamp.

In ice_init it uses pointer_file from the ice_in namelist but here it's hardcoded to 'rpointer.ice', I dont think we need both

I think either:

in ice_init you can abort with "'pointer_file' namelist not used in CESMCOUPLED,

use 'pointer_file' in ice_restart.F90 and remove the three lines in ice_init

So, our thought here is to add lpointer_file for the purposes of the multiple instances. When not in CESM:

lpointer_file = pointer_file

Does this work?

Why not just this ?:

The inst_suffix was already added in ice_init

#ifdef CESMCOUPLED write(pointer_file,'(a,i4.4,a,i2.2,a,i2.2,a,i5.5)') & pointer_file//'.',myear,'-',mmonth,'-',mday,'-',msec inquire(file=pointer_file, exist=file_exist) if (.not. file_exist) pointer_file = 'rpointer.ice'//trim(inst_suffix) #endif

Because that ends up repeating the timestamp if you are starting from a restart.

Oh got it - thanks both, adding lpointer_file sounds good.

DeniseWorthen · 2024-10-21T14:08:25Z

@dabail10 What is the current functionality (ie, why is it there?) of the stop_now variable.

CICE/cicecore/drivers/nuopc/cmeps/ice_comp_nuopc.F90

Lines 1242 to 1261 in 87c7bfa

    
           call ESMF_ClockGetAlarm(clock, alarmname='alarm_stop', alarm=alarm, rc=rc) 
        
           if (ChkErr(rc,__LINE__,u_FILE_u)) return 
        
           if (ESMF_AlarmIsRinging(alarm, rc=rc)) then 
        
              if (ChkErr(rc,__LINE__,u_FILE_u)) return 
        
              stop_now = .true. 
        
              call ESMF_AlarmRingerOff( alarm, rc=rc ) 
        
              if (ChkErr(rc,__LINE__,u_FILE_u)) return 
        
           else 
        
              stop_now = .false. 
        
           endif 
        
           call t_stopf ('cice_run_total') 
        
           ! Need to stop this at the end of every run phase in a coupled run. 
        
           call ice_timer_stop(timer_total) 
        
           if (stop_now) then 
        
              call ice_timer_print_all(stats=.true.) ! print timing information 
        
              call release_all_fileunits 
        
           endif

jedwards4b · 2024-10-21T14:18:30Z

Is there a reason that

  call ice_timer_print_all(stats=.true.) ! print timing information 
   call release_all_fileunits

are not in the ice_finalize routine? I also think that turning off the alarm at line 1248 may be problematic - why do you need to turn off a stop alarm anyway?

dabail10 · 2024-10-21T15:40:18Z

@dabail10 What is the current functionality (ie, why is it there?) of the stop_now variable.

CICE/cicecore/drivers/nuopc/cmeps/ice_comp_nuopc.F90

Lines 1242 to 1261 in 87c7bfa

call ESMF_ClockGetAlarm(clock, alarmname='alarm_stop', alarm=alarm, rc=rc)

if (ChkErr(rc,__LINE__,u_FILE_u)) return

if (ESMF_AlarmIsRinging(alarm, rc=rc)) then

if (ChkErr(rc,__LINE__,u_FILE_u)) return

stop_now = .true.

call ESMF_AlarmRingerOff( alarm, rc=rc )

if (ChkErr(rc,__LINE__,u_FILE_u)) return

else

stop_now = .false.

endif

call t_stopf ('cice_run_total')

! Need to stop this at the end of every run phase in a coupled run.

call ice_timer_stop(timer_total)

if (stop_now) then

call ice_timer_print_all(stats=.true.) ! print timing information

call release_all_fileunits

endif

I'm not sure.

eclare108213 · 2024-10-21T17:05:03Z

@dabail10 What is the current functionality (ie, why is it there?) of the stop_now variable.

I'm not sure.

stop_now is used in the standalone driver to end the run based on the value of npt.
It appears in various other drivers, which might or might not have their own methods for ending runs. Feel free to clean up your driver if it's not needed.

…for cesm

apcraig · 2024-10-23T17:07:56Z

@anton-seaice @dabail10 @DeniseWorthen

I will merge after you review and approve. This is all CMEPS driver stuff. Thanks.

DeniseWorthen

I think this is fine for UFS.

anton-seaice · 2024-10-23T21:41:20Z

There's a couple of minor review comments from me which are still open

Co-authored-by: Anton Steketee <79179784+anton-seaice@users.noreply.github.com>

…_in namelist

Wrt restart at end

dabail10 · 2024-10-25T16:33:13Z

@anton-seaice , have we addressed all of your comments?

anton-seaice

Looks good ! - Thanks

jedwards4b and others added 4 commits October 17, 2024 11:14

support for cesm write_restart_at_endofrun

56dc381

clean up merge

11da279

correct rpointer file name for cesm

b8232d7

Merge branch 'wrt_restart_at_end' of https://github.com/jedwards4b/CICE

6d8c7f6

dabail10 requested review from DeniseWorthen, anton-seaice and apcraig October 17, 2024 20:42

apcraig approved these changes Oct 17, 2024

View reviewed changes

apcraig added Drivers/coupling IO labels Oct 17, 2024

jedwards4b reviewed Oct 18, 2024

View reviewed changes

anton-seaice reviewed Oct 20, 2024

View reviewed changes

response to review

a0d1a92

correct exist and file_exist

6074f27

anton-seaice mentioned this pull request Mar 25, 2025

Name change of the pointer restart file ACCESS-NRI/access-om3-configs#415

Closed

get pointer file name root from ice_in namelist and append timestamp …

1cf1268

…for cesm

DeniseWorthen approved these changes Oct 23, 2024

View reviewed changes

dabail10 and others added 4 commits October 23, 2024 15:44

Update cicecore/cicedyn/infrastructure/io/io_pio2/ice_restart.F90

3a77e79

Co-authored-by: Anton Steketee <79179784+anton-seaice@users.noreply.github.com>

the pointer_file name is set in namelist, no need to modify here

e0b59fc

cleanup whitespace

66b4bbf

need to use default cesm pointer name for write and not read from ice…

f7552bc

…_in namelist

dabail10 added 2 commits October 24, 2024 14:03

Merge branch 'wrt_restart_at_end' into wrt_restart_at_end

8710a3a

Merge pull request #34 from jedwards4b/wrt_restart_at_end

67f96fe

Wrt restart at end

anton-seaice approved these changes Oct 28, 2024

View reviewed changes

dabail10 merged commit 7a4eb51 into CICE-Consortium:main Oct 28, 2024

NickSzapiro-NOAA mentioned this pull request Nov 19, 2024

Sync with CICE-Consortium (2024-11-18) NOAA-EMC/CICE#95

Closed

mvdebolskiy added a commit to mvdebolskiy/CICE that referenced this pull request Mar 11, 2025

Merbe (CICE-Consortium#984) 7a4eb51

1669e2e

Conversation

dabail10 commented Oct 17, 2024

PR checklist

Uh oh!

anton-seaice commented Oct 17, 2024

Uh oh!

jedwards4b commented Oct 17, 2024

Uh oh!

Choose a reason for hiding this comment

Uh oh!

anton-seaice left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DeniseWorthen commented Oct 21, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jedwards4b commented Oct 21, 2024

Uh oh!

dabail10 commented Oct 21, 2024

Uh oh!

eclare108213 commented Oct 21, 2024

Uh oh!

apcraig commented Oct 23, 2024

Uh oh!

DeniseWorthen left a comment

Choose a reason for hiding this comment

Uh oh!

anton-seaice commented Oct 23, 2024

Uh oh!

dabail10 commented Oct 25, 2024

Uh oh!

anton-seaice left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

DeniseWorthen commented Oct 21, 2024 •

edited

Loading