Opened 4 years ago

Closed 4 years ago

#1169 closed help (fixed)

Reconfiguration error - item 511 not in input dump

Reported by: mk1812 Owned by: annette
Priority: normal Component: UM Model
Keywords: Cc:
Platform: MONSooN UM Version: 8.2

Description

Hi,

I want to use the output dumps from one of my existing jobs (xizka) to provide the initial conditions for a series of other jobs, however I am getting an error whenever I run a job which requires reconfiguration of one of my xizka restart dumps. The error I get gives Error Code 30 in routine Rcf_Set_Data_Source, with the message: "Section 0 Item 511: Required field is not in input dump!". See for example my job xizkg, with output file: /home/mkasoa/xizkg000.xizkg.d13317.t180057.rcf.leave

This job is essentially an exact copy of my xizka job, just with the initial start dump changed to one of my xizka restart files - I have made no changes to any of the prognostics or user STASHmaster files, and so as far as I know there should not be any difference in the input fields needed. Moreover, if I inspect the xizka restart file (/nerc/ukca/mkasoa/xizka/atmos/restart_dumps/xizkaa.da20100901_00), then the dump actually does contain item 511 - it's field number 182 in the restart file.

The job runs successfully if I match the starting year and do not reconfigure the start dump, so clearly the dump does contain the correct fields. However some of the runs I am intending to perform will require reconfiguration of the dumps, which is why this error is problematic.

I would really appreciate any help you can give as to why I am getting this error even though the start dump apparently does contain the required field.

Best regards,
Matt Kasoar

Change History (7)

comment:1 Changed 4 years ago by annette

  • Owner changed from um_support to annette
  • Status changed from new to assigned

Hi Matt,

Your error appears on the Met Office collaboration Twiki here:
http://collab.metoffice.gov.uk/twiki/bin/view/Support/UnifiedModelErrors#Reconfiguration errors

If you don't have access to this, then you should definitely register. The solution they suggest is adding a user-STASH master file to make sure the field is set:

Add the relevant user-STASHmaster file, which will then create an entry for the prognostic in the user-Prognostics window. Either point to an ancillary or start dump that includes the prognostic, or set to zero/missing data.

It is strange that the reconfiguration doesn't find the field when it is in the dump though. Let me have a look at the job and see if I can work out what is going on.

Annette

comment:2 Changed 4 years ago by mk1812

Hi Annette,

Thanks very much for looking into this. I had seen the Collaboration Twiki entry, but I wasn't sure whether it was the same as what's happening here, since the explanation it gives for the error is that it occurs when the required prognostic is not in the input dump. What's also odd is that my original xizka job uses the reconfiguration itself, with the input dump /projects/ocean/hadgem3/initial/atmos/N96L85/akdywa.daj0910, and that didn't give any issues even though it has the same user-STASHmaster files.

I'll try implementing the solution in the Collaboration Twiki though and see if that resolves it - would you know where I could find a user-STASHmaster file that includes this field? I've had a quick check through my existing jobs and none of them have item 511 as a user prognostic at the moment.

Cheers,
Matt

comment:3 Changed 4 years ago by annette

Hi Matt,

We may be able to force it to read the field from the dump using a stash-master file but it's quite odd that the reconfiguration doesn't read in field 511.

In the reconfiguration output (from the .rcf.leave) it lists the prognostic fields in the input dump but seems to skip over field 511:

Input dump : /nerc/ukca/mkasoa/xizka/atmos/restart_dumps/xizkaa.da20100901_00
 
  ...

Input grid: 

  ...

    0    1   27840   1     0   510  4361  MEAN LAND ALBEDO AFTER TS           
    0    2   27840   1    26     4  4362  RIVER OUTFLOW       KG/M2/S     

However, if you browse the file independently with uminfo you can see that field number '4362' is stash code 511.

Leave it with me for a bit and I will ask around.

Annette

comment:4 Changed 4 years ago by mk1812

Hi Annette,

Sure thing. Thanks for your time with this - I really appreciate it!

Best,
Matt

comment:5 Changed 4 years ago by annette

Hi Matt,

I've made a user STASH-master file to get around the problem:

/home/annette/hadgem3/preSTASHmaster/inlandBasinFlow_fix_8.2

All this includes is a copy of the definition for field 511 from the global STASH master file (if you browse the .rcf.leave file you'll find it is /projects/um1/vn8.2/ctldata/STASHmaster/STASHmaster_A).

1) Add the stash master file to your job in Atmosphere → STASH → User-STASHmaster files.

2) Then go to Initialisation of user prognostics, find STASH numnber 511 and add Option code 7 and the full path to the dump file.

This works for me (see xhnkc). Hopefully this will get you going for now.

I'll leave this as an open issue though to see if I can figure out why the reconfiguration doesn't read the field in the first place.

Best wishes,
Annette

comment:6 Changed 4 years ago by mk1812

Hi Annette,

That works for me too, thanks very much for that!

As you say it would be interesting to find out what was causing this problem in the first place, but hopefully it won't be causing me any more difficulties now anyway.

Cheers,
Matt

comment:7 Changed 4 years ago by ros

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