update MOM6 to GFDL 20210308 main branch commit #458
Conversation
…rom framework directory to infra directory (2) modify CMakeLists.txt to reflect the directories being changed where *.h files are stored
|
the changes in MOM6 directories are: |
BrianCurtis-NOAA
added
enhancement
junwang-noaa
added
the
Waiting for Reviews
|
@DomHeinzeller there is no any code content change in MOM for this PR, but there are some directories structure changes: So the only stuff being touched in this PR are two MOM6 cmake files |
Thanks for the information! |
|
@dom: can you trigged the run on GAEA ? I have already launched on hera, orion, dell-p3 and cray |
|
Let's try auto-rt ... I'll add the label. |
thanks. |
|
hera and orion are fully saturated at this moment, let's hope all jobs can be done tomorrow morning. |
|
@BrianCurtis-NOAA Is there a reason the auto-rt label hasn't been added to the other platforms (jet, cheyenne) ? |
The automatic regression tests passed on gaea, log has been committed to your PR. You will need to pull those updated into your local branch if you are planning to commit more regression test logs manually. |
|
@BrianCurtis-NOAA I don't understand what you posted about jet. But I realize now that the new baseline from inlinepost first needs to be created on cheyenne before we can run against that. @DomHeinzeller do you want me to do that or have you started it already? |
|
@DeniseWorthen I do not see the code changes themselves would change the
compile time. But I do see the fluctuation of compile time from previous
tests:
***@***.*** FV3_RT]$ grep -i "elapsed=" rt_92944/*/err | more
rt_92944/compile_1/err:+ elapsed=993
***@***.*** FV3_RT]$ grep -i "elapsed=" rt_244824/*/err | more
rt_244824/compile_1/err:+ elapsed=520
***@***.*** rt_295210]$ grep -i "elapsed=" */err | more
compile_1/err:+ elapsed=1285
…On Fri, Mar 12, 2021 at 1:23 PM jiandewang ***@***.***> wrote:
@junwang-noaa <https://github.com/junwang-noaa> Would there have been
anything in the inline post PR that created longer compile times? It seems
odd.
I just had 3 failed compiling on cray
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#458 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AI7D6TKGLEE4AKDRTOEVA3DTDJE2BANCNFSM4Y2Y3RKA>
.
|
|
@junwang-noaa I tried twice on cray but all had compiling time out, 1st try had 5 cases, 2nd try had 3. How do I increase the compiling CPU time ? |
|
@jiandewang I had compile time failures for the cheyenne intel jobs also (26). I will try the auto-rt tonight and see if it is just a cheyenne congestion issue. |
|
@jaindewang. I suspect the cmake changes in your PR further may slow down
the compilation, where is your RT run directory on cray? You can change the
line 384 in ufs-weather-model/tests/rt_utils.sh
BUILD_WALLTIME="00:30:00"
to:
BUILD_WALLTIME="01:00:00"
…On Fri, Mar 12, 2021 at 3:47 PM jiandewang ***@***.***> wrote:
@junwang-noaa <https://github.com/junwang-noaa> I tried twice on cray but
all had compiling time out, 1st try had 5 cases, 2nd try had 3. How do I
increase the compiling CPU time ?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#458 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AI7D6TN7F54GUIV6MJHXRZ3TDJVTFANCNFSM4Y2Y3RKA>
.
|
@junwang-noaa see one example on cray /gpfs/hps3/stmp/Jiande.Wang/FV3_RT/rt_13574/compile_3 |
|
I am more incline to think this is related to system. For example I just show you that compile_3 had time out issue, but it finished in my 3rd try. In my previous 2 tries, the timed out cases are different. |
|
all RT passed and log files being committed, ready for review |
|
@jiandewang The MOM6 PR is approved, so go ahead and merge that and then revert the gitmodules in this PR. This PR is approved, so we can merge after that. |
|
MOM6 is merged, ufs-weather-model is pointing to dev/emc now. Thanks for everybody's work |
* upstream/develop: Updatetemplate; Fix model slowness when using threading; Update WW3 for porting and threading issues (ufs-community#383) update MOM6 to GFDL 20210308 main branch commit (ufs-community#458) Regional inlinepost (ufs-community#364) correct benchmark diag_tables for coupled model configurations; move bm_ic directory out of inputdata directory; use aws ec2 for CI test; auto-rt fixes (ufs-community#426)
## DESCRIPTION OF CHANGES: * Rename GSD_RAP13km grid to RRFS_NA_13km. * In ush/set_predef_grid_params.sh: * Change parameters of the (newly renamed) RRFS_NA_13km grid to make its write-component grid be completely within the native grid. * For the (newly renamed) RRFS_NA_13km grid, automate calculation of WRTCMP_dlon and WRTCMP_dlat from ESGgrid_DELX and ESGgrid_DELY, respectively. * NCL scripts: * Make modifications to make scripts more general and have the find_wrtcmp_grid_params.ncl script work for a write-component grid of type "rotated_latlon". * Remove unused code in NCL scripts. * Fix up comments. ## TESTS CONDUCTED: Ran the (renamed) WE2E test grid_RRFS_NA_13km on Hera. Was successful. ## CONTRIBUTORS (optional): This grid was also tested by @jaymes-kenyon on Jet.
Description
Issue(s) addressed
Closes update MOM6 to GFDL's 20210308 commit #457
MOM6 issue 60 (update MOM6 to GFDL 20210308 commit NOAA-EMC/MOM6#60)
Testing
Dependencies