pnetcdf test/DRY RUN Error in wrf simulation
BasitAli Khan
BasitAli.Khan at kaust.edu.sa
Sat Jun 7 14:17:15 CDT 2014
Hi,
Thanks for running WRF tests for this particular error. Ok, I will compile
WRF3.6 and repeat the test. I believe you have tested wrf on the same
platform that is blue gene/P. I wonder if it is possible for you to
provide your configure.wrf for wrf3.6. This will help me to ensure that I
am consistent with you on compiling wrf, netcdf, and pnetcdf.
Thanks and best regards,
b:)
----
Basit A. Khan, Ph.D.
Postdoctoral Research Fellow
Division of Physical Sciences & Engineering
Office# 3204, Level 3, Building 1,
King Abdullah University of Science & Technology
4700 King Abdullah Blvd, Box 2753, Thuwal 23955 6900,
Kingdom of Saudi Arabia.
Office: +966(0)12 808 0276, Mobile: +966(0)5 0860 3617
E-mail: basitali.khan at kaust.edu.sa
Skype name: basit.a.khan
On 6/7/14 9:07 PM, "Wei-keng Liao" <wkliao at eecs.northwestern.edu> wrote:
>Hi, Basit
>
>I just ran a WRF V3.6 test case and did not see the same error message.
>It seems that you are using WRF V3.5.1. I guess V3.6 must fix the error.
>Could you try V3.6?
>
>Wei-keng
>
>On Jun 5, 2014, at 4:00 AM, BasitAli Khan wrote:
>
>> Hi Wei-Keng and Rob,
>>
>>
>> Thanks for the prompt responses. Thanks for spotting snippet from the
>> wrong file. I get the above mentioned warnings only when i activate
>> pnetcdf that is io_form_* = 11. If i don't use pnetcdf i.e. io_form_*=2
>> the wrf sim runs perfect without any warning and/or error.
>> Sorry, not sure about "WRF_FILE_OPENED_NOT_COMMITTED" as i am not quite
>> familiar with pnetcdf stuff.
>> A detailed description of the supercomputing lab(shaheen) is also
>> available at https://www.hpc.kaust.edu.sa/documentation/user_guide/
>>
>> Following is the system info + how i compiled wrf, netCDF and pnetCDF
>>
>> - uname --all
>> Linux fen1 2.6.16.60-0.97.1-ppc64 #1 SMP Tue May 29 19:21:39 UTC 2012
>> ppc64 ppc64 ppc64 GNU/Linux
>>
>> basitali at fen1:~$ lsb_release -a
>> LSB
>> Version:
>>core-2.0-noarch:core-3.0-noarch:core-2.0-ppc:core-3.0-ppc:desktop-
>>
>>3.1-noarch:desktop-3.1-ppc32:graphics-2.0-noarch:graphics-2.0-ppc32:graph
>>ic
>> s-3.1-noarch:graphics-3.1-ppc32
>> Distributor ID: SUSE LINUX
>> Description: SUSE Linux Enterprise Server 10 (ppc)
>> Release: 10
>> Codename: n/a
>>
>>
>>
>> - The configure.wrf files is attached that describes the wrf compile
>>setup
>> - The configure statement for pnetCDF is as follows:
>> ./configure --host powerpc64-bgp-linux --build ppc64-suse-linux
>> --prefix=/home/basitali/parallel_netCDF
>> --with-mpi=/bgsys/drivers/ppcfloor/comm/xl MPICC=mpixlc MPICXX=mpixlcxx
>> MPIF77=mpixlf77 MPIF90=mpixlf90 FFLAGS="-qsuppress=cmpmsg"
>> FCFLAGS="-qsuppress=cmpmsg"
>>
>> - The configure statement for netCDF 4.1.2 is as follows:
>> ./configure --prefix=/home/basitali/netCDF_4.1.2 CC="mpixlc -qarch=450d
>> -qtune=450" CXX="mpixlcxx_r -O3 -qarch=450d -qtune=450" FC="mpixlf90_r
>>-O3
>> -qarch=450d -qtune=450" F77="mpixlf90_r -O3 -qarch=450d -qtune=450"
>> CFLAGS="-I/home/basitali/netCDF_4.1.2/include -DIBMR2Fortran
>> -I/opt/share/zlib/1.2.5/bgp-ibm/include"
>> LIBS="-L/home/basitali/netCDF_4.1.2/lib
>> -L/opt/share/zlib/1.2.5/bgp-ibm/lib -lnetcdf -lz -lm" FFLAGS="-g
>> -I/home/basitali/netCDF_4.1.2/include"
>> FCFLAGS="-I/home/basitali/netCDF_4.1.2/include -DIBMR2Fortran
>> -I/opt/share/zlib/1.2.5/bgp-ibm/include"
>> FLIBS="-L/home/basitali/netCDF_4.1.2/lib -lnetcdff
>> -L/opt/share/zlib/1.2.5/bgp-ibm/lib -lnetcdf -lz -lm" CXXFLAGS="-qpic
>> -DIBMR2Fortran -I/home/basitali/netCDF_4.1.2/include
>> -I/opt/share/zlib/1.2.5/bgp-ibm/include" CPPFLAGS="-qpic -DIBMR2Fortran
>> -I/home/basitali/netCDF_4.1.2/include
>> -I/opt/share/zlib/1.2.5/bgp-ibm/include"
>> --with-zlib=/opt/share/zlib/1.2.5/bgp-ibm --disable-netcdf-4
>> --enable-pnetcdf --with-pnetcdf=/home/basitali/parallel_netCDF
>> --disable-dap --disable-shared
>>
>>
>> uname
>> - The result of nc-config --all is as follows
>>
>> --cc -> mpixlc -qarch=450d -qtune=450
>> --cflags -> -I/home/basitali/netCDF_4.1.2/include
>> --libs -> -L/home/basitali/netCDF_4.1.2/lib -lnetcdf
>>
>> --cxx -> mpixlcxx_r -O3 -qarch=450d -qtune=450
>> --has-c++ -> yes
>>
>> --fc -> mpixlf90_r -O3 -qarch=450d -qtune=450
>> --fflags -> -g -I/home/basitali/netCDF_4.1.2/include
>> -I/home/basitali/netCDF_4.1.2/include
>> --flibs -> -L/home/basitali/netCDF_4.1.2/lib -lnetcdff -lnetcdf
>> --has-f77 -> yes
>> --has-f90 -> yes
>>
>> --has-dap -> no
>> --has-nc2 -> yes
>> --has-nc4 -> no
>> --has-hdf5 -> no
>> --has-hdf4 -> no
>> --has-pnetcdf-> yes
>> --has-szlib -> no
>>
>> --prefix -> /home/basitali/netCDF_4.1.2
>> --includedir-> /home/basitali/netCDF_4.1.2/include
>> --version -> netCDF 4.1.2
>>
>>
>>
>> Please have a look. I would greatly appreciate your help and advise in
>> this regard.
>>
>> Best regards,
>>
>>
>> b:)
>> ----
>> Basit A. Khan, Ph.D.
>>
>> Postdoctoral Research Fellow
>>
>> Division of Physical Sciences & Engineering
>>
>> Office# 3204, Level 3, Building 1,
>>
>> King Abdullah University of Science & Technology
>>
>> 4700 King Abdullah Blvd, Box 2753, Thuwal 23955 6900,
>>
>> Kingdom of Saudi Arabia.
>>
>>
>>
>> Office: +966(0)12 808 0276, Mobile: +966(0)5 0860 3617
>>
>> E-mail: basitali.khan at kaust.edu.sa
>>
>>
>>
>>
>> Skype name: basit.a.khan
>>
>>
>>
>>
>> On 6/4/14 6:54 PM, "Wei-keng Liao" <wkliao at eecs.northwestern.edu> wrote:
>>
>>> Hi, Basit
>>>
>>> The file wrf_io.F90 should be the one in external/io_pnetcdf.
>>> So the code fragment is like
>>>
>>> 2399 elseif(DH%FileStatus == WRF_FILE_OPENED_NOT_COMMITTED) then
>>> 2400 do NVar=1,MaxVars
>>> 2401 if(DH%VarNames(NVar) == VarName ) then
>>> 2402 Status = WRF_WARN_2DRYRUNS_1VARIABLE
>>> 2403 write(msg,*) 'Warning 2 DRYRUNS 1 VARIABLE
>>> (',TRIM(VarName),') in ',__FILE__,', line', __LINE__
>>> 2404 call wrf_debug ( WARN , TRIM(msg))
>>> 2405 return
>>> 2406 elseif(DH%VarNames(NVar) == NO_NAME) then
>>>
>>> It looks to me is the file is yet to be closed.
>>>
>>> Could you let us know how you build WRF (commands) and the test case
>>>(and
>>> parameters)
>>> that you run? I may give it a try to see if I can repeat your
>>>experiment
>>> and observe the same errors.
>>>
>>>
>>> Wei-keng
>>>
>>> On Jun 4, 2014, at 10:09 AM, BasitAli Khan wrote:
>>>
>>>> Hi Rob,
>>>>
>>>> Following is the snippet of the subroutine in wrf_io.F90. Actually
>>>>line
>>>> 2403 is just a variable declaration.
>>>>
>>>>
>>>> 2380 subroutine
>>>> ext_ncd_write_field(DataHandle,DateStr,Var,Field,FieldTypeIn, &
>>>> 2381 Comm, IOComm, DomainDesc, MemoryOrdIn, Stagger, DimNames,
>>>> &
>>>> 2382
>>>>
>>>>DomainStart,DomainEnd,MemoryStart,MemoryEnd,PatchStart,PatchEnd,Status)
>>>> 2383 use wrf_data
>>>> 2384 use ext_ncd_support_routines
>>>> 2385 implicit none
>>>> 2386 include 'wrf_status_codes.h'
>>>> 2387 include 'netcdf.inc'
>>>> 2388 integer ,intent(in) :: DataHandle
>>>> 2389 character*(*) ,intent(in) :: DateStr
>>>> 2390 character*(*) ,intent(in) :: Var
>>>> 2391 integer ,intent(inout) :: Field(*)
>>>> 2392 integer ,intent(in) :: FieldTypeIn
>>>> 2393 integer ,intent(inout) :: Comm
>>>> 2394 integer ,intent(inout) :: IOComm
>>>> 2395 integer ,intent(in) :: DomainDesc
>>>> 2396 character*(*) ,intent(in) :: MemoryOrdIn
>>>> 2397 character*(*) ,intent(in) :: Stagger ! Dummy
>>>> for
>>>> now
>>>> 2398 character*(*) ,dimension(*) ,intent(in) :: DimNames
>>>> 2399 integer ,dimension(*) ,intent(in) :: DomainStart,
>>>> DomainEnd
>>>> 2400 integer ,dimension(*) ,intent(in) :: MemoryStart,
>>>> MemoryEnd
>>>> 2401 integer ,dimension(*) ,intent(in) :: PatchStart,
>>>> PatchEnd
>>>> 2402 integer ,intent(out) :: Status
>>>> 2403 integer :: FieldType
>>>> 2404 character (3) :: MemoryOrder
>>>> 2405 type(wrf_data_handle) ,pointer :: DH
>>>> 2406 integer :: NCID
>>>> 2407 integer :: NDim
>>>> 2408 character (VarNameLen) :: VarName
>>>> 2409 character (3) :: MemO
>>>> 2410 character (3) :: UCMemO
>>>> 2411 integer :: VarID
>>>> 2412 integer ,dimension(NVarDims) :: Length
>>>> 2413 integer ,dimension(NVarDims) :: VDimIDs
>>>> 2414 character(80),dimension(NVarDims) :: RODimNames
>>>> 2415 integer ,dimension(NVarDims) :: StoredStart
>>>> 2416 integer ,dimension(:,:,:,:),allocatable :: XField
>>>> 2417 integer :: stat
>>>> 2418 integer :: NVar
>>>> 2419 integer :: i,j
>>>> 2420 integer ::
>>>>i1,i2,j1,j2,k1,k2
>>>> 2421 integer ::
>>>>x1,x2,y1,y2,z1,z2
>>>> 2422 integer ::
>>>>l1,l2,m1,m2,n1,n2
>>>> 2423 integer :: XType
>>>> 2424 integer :: di
>>>> 2425 character (80) :: NullName
>>>> 2426 logical :: NotFound
>>>> 2427
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> Cheers,
>>>> b:)
>>>> ----
>>>> Basit A. Khan, Ph.D.
>>>>
>>>> Postdoctoral Research Fellow
>>>>
>>>> Division of Physical Sciences & Engineering
>>>>
>>>> Office# 3204, Level 3, Building 1,
>>>>
>>>> King Abdullah University of Science & Technology
>>>>
>>>> 4700 King Abdullah Blvd, Box 2753, Thuwal 23955 6900,
>>>>
>>>> Kingdom of Saudi Arabia.
>>>>
>>>>
>>>>
>>>> Office: +966(0)12 808 0276, Mobile: +966(0)5 0860 3617
>>>>
>>>> E-mail: basitali.khan at kaust.edu.sa
>>>>
>>>>
>>>>
>>>>
>>>> Skype name: basit.a.khan
>>>>
>>>>
>>>>
>>>>
>>>> On 6/4/14 4:58 PM, "Rob Latham" <robl at mcs.anl.gov> wrote:
>>>>
>>>>>
>>>>>
>>>>> On 06/04/2014 01:55 AM, BasitAli Khan wrote:
>>>>>> Hi there,
>>>>>> I sent you the following email a a weak ago regarding pnetcdf make
>>>>>> check errors, but did not get response. I used this pnetcdf to
>>>>>> compile
>>>>>> netCDF library 4.1.2. The pnetcdf was successfully linked and netCDF
>>>>>> was
>>>>>> successfully compiled. I ran a WRF simulation which also was
>>>>>>completed
>>>>>> successfully, however, the log file rsl.out.0000 (attached ) has
>>>>>> several
>>>>>> (DRY RUN) errors that are originating from pnetCDF. I am wondering
>>>>>> what
>>>>>> these errors are and would they be affecting my simulation in
>>>>>>anyway?
>>>>>>
>>>>>> Thanks in advance and please do REPLY this email.
>>>>>
>>>>> Hello! sorry for the oversight. Here is my (list-only) reply from
>>>>>last
>>>>> week:
>>>>>
>>>>>
>>>>>http://lists.mcs.anl.gov/pipermail/parallel-netcdf/2014-May/001574.htm
>>>>>l
>>>>>
>>>>> Thanks for subscribing to the mailing list.
>>>>>
>>>>> There are folks using pnetcdf and WRF on Blue Gene. I don't know the
>>>>> details, but it apparently is possible. I think we can get you
>>>>> straightened out.
>>>>>
>>>>> When you see errors like this:
>>>>>
>>>>> /bin/sh: line 1: 6418 Illegal instruction ./nc_test -c -d .
>>>>>
>>>>> it means you are running back-end code on the login nodes. Blue Gene
>>>>> is
>>>>> a cross-compile environment: compute node codes cannot run on the
>>>>>login
>>>>> nodes. (well, technically they can... until they make a system call)
>>>>>
>>>>> what does line 2403 of wrf_io.f90 look like?
>>>>>
>>>>> ==rob
>>>>>
>>>>>
>>>>> Hi Rob,
>>>>
>>>>
>>>>>>
>>>>>> Best regards,
>>>>>> b:)
>>>>>> ----
>>>>>>
>>>>>> Basit A. Khan, Ph.D.
>>>>>>
>>>>>> Postdoctoral Research Fellow
>>>>>>
>>>>>> Division of Physical Sciences & Engineering
>>>>>>
>>>>>> Office# 3204, Level 3, Building 1,
>>>>>>
>>>>>> King Abdullah University of Science & Technology
>>>>>>
>>>>>> 4700 King Abdullah Blvd, Box 2753, Thuwal 23955 6900,
>>>>>>
>>>>>> Kingdom of Saudi Arabia.
>>>>>>
>>>>>> Office: +966(0)12 808 0276, Mobile: +966(0)5 0860 3617
>>>>>>
>>>>>> E-mail: basitali.khan at kaust.edu.sa
>>>>>><mailto:basitali.khan at kaust.edu.sa>
>>>>>>
>>>>>> Skype name: basit.a.khan
>>>>>>
>>>>>> From: KAUST <basitali.khan at kaust.edu.sa
>>>>>> <mailto:basitali.khan at kaust.edu.sa>>
>>>>>> Date: Wednesday, May 28, 2014 3:05 PM
>>>>>> To: "parallel-netcdf at mcs.anl.gov
>>>>>><mailto:parallel-netcdf at mcs.anl.gov>"
>>>>>> <parallel-netcdf at mcs.anl.gov <mailto:parallel-netcdf at mcs.anl.gov>>
>>>>>> Subject: pnetcdf test
>>>>>>
>>>>>> Hi there,
>>>>>> I am new to pnetcdf. I want to compile it for faster IO of WRF
>>>>>>model
>>>>>> on
>>>>>> our bluegeneP with Linux RedHat 10 on it. I apparently successfully
>>>>>> installed the latest version of pnetcdf 1.4.1, however when i run
>>>>>> 'make
>>>>>> testing', i got few errors (log file attached). I would greatly
>>>>>> appreciate if you could have a look and advise where i am going
>>>>>>wrong.
>>>>>>
>>>>>> The compilation of parallel netcdf created 4 directories. The
>>>>>>contents
>>>>>> of these dir are as follows:
>>>>>>
>>>>>> /lib/libpnetcdf.a
>>>>>> /include/pnetcdf.h pnetcdf.inc pnetcdf.mod
>>>>>> /bin/ncmpidiff ncmpidump ncmpigen ncmpivalid
>>>>>> /man/man1/ncmpidiff.1 ncmpidump.1 ncmpigen.1 ncmpivalid.1
>>>>>> /man/man3/pnetcdf.3 pnetcdf_f77.3 pnetcdf_f90.3
>>>>>>
>>>>>>
>>>>>> Another question is that, do i have to do anything special to use it
>>>>>> with WRF?
>>>>>>
>>>>>> Thanks in advance and best regards,
>>>>>>
>>>>>>
>>>>>> b:)
>>>>>> ----
>>>>>>
>>>>>> Basit A. Khan, Ph.D.
>>>>>>
>>>>>> Postdoctoral Research Fellow
>>>>>>
>>>>>> Division of Physical Sciences & Engineering
>>>>>>
>>>>>> Office# 3204, Level 3, Building 1,
>>>>>>
>>>>>> King Abdullah University of Science & Technology
>>>>>>
>>>>>> 4700 King Abdullah Blvd, Box 2753, Thuwal 23955 6900,
>>>>>>
>>>>>> Kingdom of Saudi Arabia.
>>>>>>
>>>>>> Office: +966(0)12 808 0276, Mobile: +966(0)5 0860 3617
>>>>>>
>>>>>> E-mail: basitali.khan at kaust.edu.sa
>>>>>><mailto:basitali.khan at kaust.edu.sa>
>>>>>>
>>>>>> Skype name: basit.a.khan
>>>>>>
>>>>>>
>>>>>>
>>>>>>---------------------------------------------------------------------
>>>>>>--
>>>>>> -
>>>>>>
>>>>>> This message and its contents including attachments are intended
>>>>>> solely
>>>>>> for the original recipient. If you are not the intended recipient or
>>>>>> have received this message in error, please notify me immediately
>>>>>>and
>>>>>> delete this message from your computer system. Any unauthorized use
>>>>>>or
>>>>>> distribution is prohibited. Please consider the environment before
>>>>>> printing this email.
>>>>>
>>>>> --
>>>>> Rob Latham
>>>>> Mathematics and Computer Science Division
>>>>> Argonne National Lab, IL USA
>>>>
>>>>
>>>> ________________________________
>>>>
>>>> This message and its contents including attachments are intended
>>>>solely
>>>> for the original recipient. If you are not the intended recipient or
>>>> have received this message in error, please notify me immediately and
>>>> delete this message from your computer system. Any unauthorized use or
>>>> distribution is prohibited. Please consider the environment before
>>>> printing this email.
>>>
>>
>>
>> ________________________________
>>
>> This message and its contents including attachments are intended solely
>>for the original recipient. If you are not the intended recipient or
>>have received this message in error, please notify me immediately and
>>delete this message from your computer system. Any unauthorized use or
>>distribution is prohibited. Please consider the environment before
>>printing this email.
>> <configure.wrf>
>
________________________________
This message and its contents including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.
More information about the parallel-netcdf
mailing list