#3006 closed help (fixed)

Yet another problem with soil & veg ancillaries

Reported by: charlie Owned by: um_support
Component: UM Model Keywords:
Cc: Platform: NEXCS
UM Version: 10.7

Description

Hi,

Sorry to bother you again, but I'm having more problems with some of my ancillary files, specifically soil (both soil parameters and soil dust) and vegetation. My suite is u-bm327. I know that it must be a problem with one of these 3 ancils, because a previous version of the suite (without these 3) works fine. It is only these 3 that I have modified:

/home/d05/cwilliams/pliocene/gc31/ancils/soil/dust
/home/d05/cwilliams/pliocene/gc31/ancils/soil/parameters
/home/d05/cwilliams/pliocene/gc31/ancils/vegetation/frac

I have checked all the usual schoolboy errors e.g. that the ancillaries are in the right format and UM version, the right way up, that the relevant fields (e.g. vegetation fraction) sum to 1, and the soil/veg fields are not inconsistent with each other e.g. vegetation where there is zero soil parameters, or vice versa. These are all problems I have encountered before, but everything looks fine. I have also tried swapping out one or both of the files, e.g. using the original veg with my soil files, and that also fails - implying there is possibly something wrong with all of these 3 i.e. soil parameters, soil dust and veg.

That being said, I am also running another suite (u-bk944), still using the same veg fraction file but not the soils files, and that is running fine now. So this implies my veg is okay, and there is something wrong with it's interaction with the soils, doesn't it?

The error message in my job.err is non-specific:

_pmiu_daemon(SIGCHLD): [NID 03859] [c6-1c0s4n3] [Mon Sep  2 11:07:01 2019] PE RANK 1045 exit signal Aborted
atpAppSigHandler: Back-end never delivered its pid. Re-raising signal.
[NID 03859] 2019-09-02 11:07:01 Apid 76641603: initiated application termination
[FAIL] run_model # return-code=137
2019-09-02T11:07:05Z CRITICAL - failed/EXIT

Would somebody be able to take a look at the above 3 ancils, in case I have missed something?

Many thanks,

Charlie

Change History (7)

comment:1 Changed 13 months ago by grenville

Charlie

u-bm327 failed because (see ocean.output):

==⇒>> : E R R O R

===========

iom_open ~

File ./restart_icebergs.nc* not found

Not because of bad ancillary files.

Grenville

comment:2 Changed 13 months ago by charlie

Hi Grenville,

Right, thanks. Ummmm… I don't entirely understand what that means, or what that file is? I have never seen it before. I have just checked my various restarts, and I think they are all correct - the atmosphere restart dump is being pointed to at um > namelist > Model input and output > Dumping and meaning > astart (with reconfiguration turned off in suite conf > Build and run), the ocean restart dumps are being pointed to at nemo_cice > Restart files > Nemo restart (for the physical restart) and ocean_passive_tracers > env > Initialisation settings (for the tracer file), and the CICE restart dump is being pointed to at nemo_cice > Restart files > CICE restart. Is this not correct? I thought, according to the website instructions, that this is all that was needed to restart a new suite from an existing dump? I have never seen/used a restart_icebergs.nc before. Am I missing something obvious?

Charlie

comment:3 Changed 13 months ago by grenville

Hi Charlie

The problem arises because of where you have specified the NEMO start file - you have it under nemo_cice→Restart files. If instead, you specify it under suite conf→run initialisation (and leave blank the field under nemo_cide..), then the model runs (well, it did for me).

I have not understood why this results in an error with iceberg files.

Grenville

comment:4 Changed 13 months ago by charlie

Okay, I will do that. Is that the same for the other 2 options in suite conf > Run initialisation, i.e. NEMO icebergs and CICE? Or do I leave those 2 as they are, and still specify the cice restart dump in nemo_cice > Restart files? Which overrides which?

Charlie

comment:5 Changed 13 months ago by grenville

What I did worked - you can try other combinations.

comment:6 Changed 13 months ago by charlie

Very many thanks, that appears to have worked. I just hope it keeps running now! I will close this ticket either way, because if it fails it will be due to a different problem.

Thanks again,

Charlie

comment:7 Changed 13 months ago by charlie

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.