Unable to pass all the tests with pnetcdf 1.6.1, Intel 15.0.3.048 and Mvapich2 2.1

Wei-keng Liao wkliao at eecs.northwestern.edu
Fri Sep 18 17:15:41 CDT 2015


Hi, Craig

Could you send us your config.log file?

Wei-keng

On Sep 18, 2015, at 4:57 PM, Craig Tierney - NOAA Affiliate wrote:

> Hello,
> 
> I am trying to build Pnetcdf 1.6.1 with Intel 15.0.3.048 and Mvapich2 2.1.  I am getting some failures in the tests.  I so far am getting 4 failures.  The flash-io test fails and the make process exits.  When I build Pnetcdf 1.6.1 with Intel 15.0.3.048 and Intel MPI 5.1.1.109, all tests complete cleanly.
> 
> Are there known issues with Pnetcdf 1.6.1 and Mvapich2 2.1?
> 
> Thanks,
> Craig
> 
> ---------------
> ./nc_test -d       /lfs3/jetmgmt/Craig.Tierney
> 
> 	FAILURE at line 112 of test_read.c: ncmpi_open of non-netCDF file: status = -220
> 	### 1 FAILURES TESTING test_ncmpi_open! Stop ... ###
> 
> ./nc_test: expects 0 failures ... Total number of failures: 1
> *** TESTING C   ./nc_test for format CDF-1                         ------ failed
> 
> --------------
> And:
> 
> -------------
> *** TESTING C   ./varn_int for ncmpi_put_varn_int_all()            ------ pass
> Error: expect error code -220 but got -51
> Error: expect error code -220 but got -51
> Error: expect error code -220 but got -51
> Error: expect error code -220 but got -51
> *** TESTING C   ./modes for file create/open modes                 ------ failed
> *** TESTING C   ./one_record for only one record variable          ------ pass
> 
> ----------------
> 
> Also:
> 
> ---------------
> Error at line=71: NetCDF: Not a valid ID
> Error at line=72: NetCDF: Not a valid ID
> Error at line=69: NetCDF: Not a valid ID
> Error at line=71: NetCDF: Not a valid ID
> Error at line=72: NetCDF: Not a valid ID
> *** TESTING C   ./wait_after_indep for ncmpi_end_indep_data        ------ fail
> 
> ----------
> Lastly:
> 
> make -w -C FLASH-IO ptest
> make[2]: Entering directory `/misc/whome/admin/software/pnetcdf/parallel-netcdf-1.6.1-intel-mvapich2/benchmarks/FLASH-IO'
> mpiexec.hydra -np 4 ./flash_benchmark_io /lfs3/jetmgmt/Craig.Tierney/flash_io_test_
>  nfmpi_enddef
>  (Internal error) beginning file offset of this variable is inconsistent among p
>  r
>  nfmpi_enddef
>  (Internal error) beginning file offset of this variable is inconsistent among p
>  r
>  nfmpi_enddef
>  (Internal error) beginning file offset of this variable is inconsistent among p
>  r
> [cli_1]: aborting job:
> application called MPI_Abort(MPI_COMM_WORLD, -1) - process 1
> [cli_2]: aborting job:
> application called MPI_Abort(MPI_COMM_WORLD, -1) - process 2
> [cli_3]: aborting job:
> application called MPI_Abort(MPI_COMM_WORLD, -1) - process 3
> 
> ===================================================================================
> =   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
> =   PID 9284 RUNNING AT fe1
> =   EXIT CODE: 255
> =   CLEANING UP REMAINING PROCESSES
> =   YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
> ===================================================================================
> make[2]: *** [ptest] Error 255



More information about the parallel-netcdf mailing list