Restart problems when running CESM1.0.4 with 0.23 degree

[weitstudy@126_com]

Deal all,I am runing CESM1.0.4 with F1850_CAM5 compset and f02_f02 resolution in the AIX operation system. I run the model for five days and then change 'start_type = 'continue'' in drv_in script to restrat the model. But the restart is failing. The error message is as following.  Can you help me? Thank you very much!---------------------------------------------------------------------- 511:(shr_stream_set) filename = /cma/g5/weiting/model/inputdata/atm/cam/sst/sst_HadOIBl_bc_0.23x0.31_clim_pi_c101028.nc 511:(seq_frac_check) [lnd init] afrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [lnd init] lfrac min/max   =   0.143709186256146815E-02   1.00000000000000000 511:(seq_frac_check) [lnd init] lfrin min/max   =   0.143709186256146815E-02   1.00000000000000000 511:(seq_frac_check) [ice init] afrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [ice init] ofrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [ice init] ifrac min/max   =    0.00000000000000000       0.00000000000000000 511:(seq_frac_check) [ocn init] afrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [ocn init] ofrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [ocn init] ifrac min/max   =    0.00000000000000000       0.00000000000000000 511:(seq_frac_check) [atm init] afrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [atm init] lfrac min/max   =    0.00000000000000000       0.00000000000000000 511:(seq_frac_check) [atm init] ofrac min/max   =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [atm init] ifrac min/max   =    0.00000000000000000       0.00000000000000000 511:(seq_frac_check) [atm init] lfrin min/max   =    0.00000000000000000       0.00000000000000000 511:(seq_frac_check) [atm init] sum min/max     =    1.00000000000000000       1.00000000000000000 511:(seq_frac_check) [atm init] sum ncnt/maxerr =        0   0.00000000000000000   1: WARNING:  water balance error  nstep =  481  indexc=  1710  errh2o=  15.0576070139895410   1: clm urban model is stopping - error is greater than 1.e-1   1: nstep =  481  indexc=  1710  errh2o=  15.0576070139895410   1: ctype(indexc):  65   1: forc_rain    =  0.000000000000000000E+00   1: forc_snow    =  0.757670546070757430E-07   1: endwb        =  43623.7341441084573   1: begwb        =  43608.6669105444744   1: qflx_evap_tot=  -0.106203996046594954E-04   1: qflx_irrig   =  0.000000000000000000E+00   1: qflx_surf    =  0.000000000000000000E+00   1: qflx_qrgwl   =  0.000000000000000000E+00   1: qflx_drain   =  0.000000000000000000E+00 Write failed: Broken pipe0.000000000000000000E+00----------------------------------------------------------------------

 

[wanlingfeng_123@163_com]

also you need change the restart date.

 

[santos]

Don't change "start_type" directly. Instead change CONTINUE_RUN to TRUE in the env files. This can be done with the following command in your case directory:./xmlchange -file env_run.xml -id CONTINUE_RUN -val TRUE

 

[weitstudy@126_com]

Thank you! But it still doesn't work. 

 

[weitstudy@126_com]

santos, thank you very much . But the problem is still not resolved. For the low resolution compset, it works well. But for the high resolution one, the restart is bad. 

 

[santos]

The "broken pipe" issue may be due to running out of wall clock time, or because some communication operation has timed out.I don't have much experience with high resolution runs, though, so aside from making sure that the system is giving you enough time, I'm not sure what to do.

 

[w_xuej@163_com]

I change CONTINUE_RUN to TRUE in the env files.But I restart from the beginning .Why?