Opened 6 years ago

Closed 6 years ago

#1196 closed help (fixed)

run keeps failing because quota exceeded

Reported by: s1251469 Owned by: grenville
Component: UM Model Keywords:
Cc: Platform: HECToR
UM Version: 7.3

Description

Hello,

My job xjhxc keeps failing because quota is exceeded. It is a standard release job for 1 year outputting some variables as monthly averages. I don't know why that is exceeding my quota.

The biggest archived files are actually the upmean and some other streams I don't know (…da… and …ps…). Maybe you tell me how to stop these being archived?

I also don't know how to change the start date so that I can just get the months not already done. I tried changing the
model selection → input/output… → start date and run length….
previously but that created some sort of error. Is there a change to make somewhere else?

I also have a separate question. I would like to output cloud ice and graupel prognostic from section 0 as monthly means but the "verify diagnostics" tells me they are not available in this configuration. What do I need to do to archive these?

Change History (6)

comment:1 Changed 6 years ago by s1251469

Just to let you know that I have managed to complete this run by deleting some other files.

I am still interested in the answers to my questions though:
1) Can the archiving of da/ps/pm streams be stopped or are they essential to restarting?
2) Is there more to changing the start date that the obvious?
3) can I get monthly means of ice crystals (and graupel) from section 0 archived?

many thanks,
Declan

comment:2 Changed 6 years ago by grenville

Declan

I increased your quota on HECToR.

I am not sure what you have achieved by archiving - all that's happening is that data is being moved from where it is written ($DATAW) to ($DATAM/archive).

The da files are start files, which may be useful for rerunning the model, however, with climate meaning switched on, many other files appear which should be kept if you want the run to continue properly. Do you really want climate meaning in this run?

To change the start date, you will need an appropriate start file and the correct updating ancillary files.

Your model can't produce graupel or ice crystals - you need to switch on the second cloud ice prognostic and graupel in model selection→scientific parameters→section by section choices→section4 large scale precip. However, your start files may need to include these as prognostics before the run can start so you will need to reconfigure them in. In addition, there may be branches required to make these options work properly - best to consult with an expert before simply switching them on.

Grenville

comment:3 Changed 6 years ago by grenville

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

comment:4 Changed 6 years ago by s1251469

Thanks Grenville,

That's really helpful.

I am just running a standard release job from UKCA and trying to understand what's going on. I assume you can't provide any more enlightenment as to what the purpose of archiving or climate meaning is? As this may help me understand why it is happening. If you're not sure then don't worry, I can ask Luke Abraham at some point if it becomes important.

How is climate meaning switched off? By removing all the upmean diagnostics from stash?

regards,
Declan

comment:5 Changed 6 years ago by grenville

Declan

Typically you'd use archiving to put important data somewhere other than /work (the fast parallel file system which is really only for temporary storage on ARCHER for example) - on MONSooN you might move data to the MASS archive, on HECToR (ARCHER) to the RDF.

I'd switch off archiving for now.

Climate meaning is just a clever way to create means over long time periods. You can switch it off in model selection→atmosphere→control→post processing dumping and meaning→dumping and meaning and deselect Defining a meaning sequence.

Grenville

comment:6 Changed 6 years ago by grenville

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