Main menu

Navigation

-res T62_g16 -compset G doesn't work?

9 posts / 0 new
Last post
xinmiao@...
-res T62_g16 -compset G doesn't work?

Hi,

I have successfully compiled and run '-res T62_g37 -compset G' in CESM1.1.1.

However, when I used the same set of parameters/ configurations but different resolution ''-res T62_g16 -compset G', the model can be compiled without a warning but failed to excute. I checked the log files, in cpl.log:

cpl.log.130516-131514

(seq_mct_drv) : Initialize each component: atm, lnd, rof, ocn, ice, and glc

(seq_mct_drv) : Initialize atm component ATM

(seq_mct_drv) : Initialize lnd component LND

(seq_mct_drv) : Initialize rof component ROF

(seq_mct_drv) : Initialize ocn component OCN

(seq_mct_drv) : Initialize ice component ICE

(shr_sys_abort) ERROR: restformat: number of records on restart file not supported

(shr_sys_abort) WARNING: calling shr_mpi_abort() and stopping

What's the meaning of "number of records on restart file not supported"? How to fix it? What's the possible problem? Thanks!

------------------------------------------

Also I listed the error messages in ccsm.log file:

ccsm.log.130516-131514

CalcWorkPerBlock: Total blocks:   256 Ice blocks:    82 IceFree blocks:   160 Land blocks:    14

[compute-0-6:11271] *** Process received signal ***

[compute-0-6:11271] Signal: Segmentation fault (11)

[compute-0-6:11271] Signal code: Address not mapped (1)

[compute-0-6:11271] Failing at address: (nil)

[compute-0-6:11271] [ 0] /lib64/libpthread.so.0 [0x38bf80e7c0]

[compute-0-6:11271] [ 1] /share/apps/openmpi/1.4.3-gcc-4.6.1/lib/libmpi_f77.so.0(MPI_ABORT+0x1b) [0x2affd2dd12ab]

[compute-0-6:11271] [ 2] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__shr_mpi_mod_MOD_shr_mpi_abort+0xd1) [0x87eaf1]

[compute-0-6:11271] [ 3] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__shr_sys_mod_MOD_shr_sys_abort+0x210) [0x8beb66]

[compute-0-6:11271] [ 4] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__ice_exit_MOD_abort_ice+0x10) [0x50a060]

[compute-0-6:11271] [ 5] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__ice_restart_MOD_restformat+0x168) [0x54997c]

[compute-0-6:11271] [ 6] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__ice_restart_MOD_restartfile_bin+0x57c) [0x54fa3a]

[compute-0-6:11271] [ 7] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__cice_initmod_MOD_cice_init+0x165) [0x588605]

[compute-0-6:11271] [ 8] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__ice_comp_mct_MOD_ice_init_mct+0x21a6) [0x4f2ff0]

[compute-0-6:11271] [ 9] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(__ccsm_comp_mod_MOD_ccsm_init+0x80ba) [0x485e72]

[compute-0-6:11271] [10] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe(main+0x4a) [0x49518a]

[compute-0-6:11271] [11] /lib64/libc.so.6(__libc_start_main+0xf4) [0x38bec1d994]

[compute-0-6:11271] [12] /home/xin.miao/CESM/cesm1_1_1/scripts/testG16/EXE/cesm.exe [0x470799]

[compute-0-6:11271] *** End of error message ***

--------------------------------------------------------------------------

mpirun noticed that process rank 0 with PID 11271 on node compute-0-6.local exited on signal 11 (Segmentation fault).

--------------------------------------------------------------------------

 

 

Shane

dbailey

This should really go in the CICE modeling section. This means you are trying to read an initial/restart file in the ice that is not supported. What type of run is this? Initial/hybrid/branch .. etc? What is ice_ic set to in your ice_in file?

xinmiao@...

Hi, dbailey,

I confirm that it is an initial run: in env_run.xml -- <entry id="RUN_TYPE"   value="startup"  />    This is the default setting, I didn't change it.

In ice_in file, ice_ic = '/home/xin.miao/CESM/inputdata/ice/cice/iced.0001-01-01.gx1v6_20080212'

This file was manually downloaded, because the previous '-res T62_g37 -compset G' version can't download it automatically, -- please refer to

http://bb.cgd.ucar.edu/weird-problem-compset-h-running

So I guess I need another ice_in restart file after I changed the resolution? It did not give me any warning when compiling.

BTW, what's the function of 'iced.0001-01-01.gx1v6_20080212' anyway? Why do I need it for restarting?

Thanks, Shane

Shane

dbailey

Hi Shane,

The CICE needs an initial or restart file to start the state from just as any of the components do. I noticed that your ice_ic is pointing to a gx1v6 restart, but your resolution is T62_g37, i.e. the ocean and ice are on the gx3v7 or nominal 3-degree grid. You must have changed the ice_ic namelist value? Try setting up a new case and see what restart file it points you to. Then you'll need to manually download that one. I have a feeling something was lost in the previous communication and you inadvertently downloaded the gx1v6 restart.

Dave

xinmiao@...
Dave, thanks for your hint, I figure it out -- yes, it is still the ice_in file problem. Last time in T62_g37 case, the model failed to download iced.0001-01-01.gx3v7_20080212 and give me a warning. This time when I tried T62_g16,it seems successfully download 'iced.0001-01-01.gx1v6_20080212' file but for some reasons, the file size is 0! And there's no warning! After manually download the corresponding initial ice file, the problem solved. I deeply appreciate your quick response! Shane

Shane

ljdiao@...

Hi, Xinmiao:


I am struggling on installing CESM1.1.1 on my machine. Can I ask which compiler you used for installing CESM1.1.1, intel or PGI? I am using PGI and got problems with that. thanks,

Lijun Diao

xinmiao@...

Lijun,

I guess your are talking about using intel or PGI fortran to compile the model? Actually, I understand CESM model should be comiled/ build in Linux based computers/ clusters, and what you need to do is to follow the instruction

http://www.cesm.ucar.edu/models/cesm1.1/cesm/doc/usersguide_new/book1.html

You really don't need to compile by yourself. Give me your email, I can send you my script for your reference. Shane

 

 

Shane

ljdiao@...

Hi, Xin Miao

Thanks for your providing help. The online user manual looks better than the pdf one.

I compiled CESM1.1.1 on our linux cluster and got a series of problems seemingly related to the compiler or path configuration. So I am wondering which compiler you are using and if possible its version.

thanks in advance for telling me that and please don't forget to send me your script at ljdiao@gmail.com

Lijun Diao

j08lue@...

For the initial problem (regarding CICE input files), see this post for a potential solution:

https://bb.cgd.ucar.edu/port-validation-runtime-failure-error-restformat...

Log in or register to post comments

Who's new

  • m.kliphuis@...
  • ddc3061993@...
  • hui.ding@...
  • zhouc@...
  • arianna.valmass...