merged_ptype issues in restart #659
Closed
PerryShafran-NOAA
started this conversation in
Issues
Replies: 2 comments 3 replies
-
I won't be able to help until after the ecFlow training, but please update here on any progress you make. |
Beta Was this translation helpful? Give feedback.
3 replies
-
We can close this issue. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello, everyone,
Restart is by and large working as it should for restart except for ptype, which I don't understand how this can be fixed.
In a full interrupted run, there is no error for ptype. Only in restart cases do I see this error. It begins with a WARNING that certain files are missing:
Later, we see an error in the METplus run:
01/29 14:04:04.733 metplus.bd12844c (command_builder.py:264) ERROR: (command_builder.py:619) Could not find FCST_INPUT file /lfs/h2/emc/stmp/perry.shafran/evs_test/prod/tmp/jevs_mesoscale_nam_grid2obs_stats_00.178317615.cbqs01/grid2obs/data/nam/merged_ptype/merged_ptype_metar_conusp_job361_init2025012800_fhr000.nc using template /lfs/h2/emc/stmp/perry.shafran/evs_test/prod/tmp/jevs_mesoscale_nam_grid2obs_stats_00.178317615.cbqs01/grid2obs/data/nam/merged_ptype/merged_ptype_metar_conusp_job361_init{init?fmt=%Y%m%d%H}_fhr{lead?fmt=%3H}.nc
Why is METplus being run in this case? I'm not sure. The way the code is written now, if the job is written in the completed_jobs.txt file, then everything should be commented out, including any calls to METplus.
But when I look at the job361 job file in the restart run, I see this:
You can see that there are some jobs that are not commented out. Jobs are supposed to be commented out (thus not run) when the item exists in the completed_jobs.txt file (and the job is in the completed text file - 6 times actually).
It should also be noted that at the start of the code, a whole bunch of PTYPE stat files from the restart directory are copied over to the working directory, so it's clear that ptype data had been done in the interrupted run and copied over, thus no need to do any more ptype processing. The final stat file for the restart job is identical to the final stat file of the full, uninterrupted job. The question is why any of the uncommented jobs are listed in the job file - that would only be the case if job361 didn't exist in the completed_jobs.txt file.
@MarcelCaron-NOAA @malloryprow
Beta Was this translation helpful? Give feedback.
All reactions