[GEMPAK #MCL-633919]: gpend not clearing resources?

Hi Nancy,

Each instance of os.system in python will execute the given command in a 
separate subshell, so while gdgrib is running in one, gpend will run in another 
and will not be able to close the gplt process created by gdgrib (similar to 
closing a terminal window before running gpend).  My python isn't up to speed, 
but it appears you should be able to execute multiple commands within the same 
subshell to solve this problem.  

This page seems to provide a thorough rundown of python system interfaces.

http://docs.python.org/library/os.html

Best,

Michael James
Unidata


> Full Name: Nancy Rehak
> Email Address: address@hidden
> Organization: NCAR/RAL
> Package Version: 5.11.4
> Operating System: Linux Debian 5.0.4
> Hardware:
> Description of problem: I'm using the GEMPAK gdgrib utility to convert GEMPAK 
> files received from another organization into GRIB1 format.  We are receiving 
> the files in realtime so I'm using a Python script to call the gdgrib 
> utility.  I also call gpend after each gdgrib call to try to clear resources. 
>  However, after my Python script runs, I still see message queues and gplt 
> processes.  Here is what I'm doing in my Python script:
> 
> #
> # Write the GEMPAK commands to a temporary file
> #
> 
> pid = os.getpid()
> gfile_path = '/tmp/gempak_params.%d' % pid
> print "pid = ", pid
> print "gfile_path = ", gfile_path
> sys.stdout.flush()
> 
> gfile = open(gfile_path, 'w')
> 
> gfile.write("GDFILE=" + input_path + "\n")
> gfile.write("GFUNC=" + gfunc + "\n")
> gfile.write("GDATTIM=" + fcst + "\n")
> gfile.write("GLEVEL=0\n")
> gfile.write("GVCORD=NONE\n")
> gfile.write("GBTBLS=" + table + "\n")
> gfile.write("GBFILE=" + output_path + "\n")
> gfile.write("VERCEN=" + vercen + "\n")
> gfile.write("PDSVAL=" + pdsval + "\n")
> gfile.write("PRECSN=" + precsn + "\n")
> gfile.write("WMOHDR=\n")
> gfile.write("CPYFIL=\n")
> gfile.write("PROJ=\n")
> gfile.write("GRDAREA=\n")
> gfile.write("KXKY=\n")
> gfile.write("RUN\n")
> 
> gfile.close()
> 
> #
> # Run the GEMPAK command
> #
> 
> print "About to run gdgrib"
> sys.stdout.flush()
> 
> command = "gdgrib < " + gfile_path
> os.system(command)
> 
> print "Ran gdgrib"
> sys.stdout.flush()
> 
> # End the GEMPAK process
> 
> print "About to run gpend"
> sys.stdout.flush()
> 
> command = "gpend"
> os.system(command)
> 
> print "Ran gpend"
> sys.stdout.flush()
> 
> #
> # Delete the temporary file
> #
> 
> os.remove(gfile_path)
> 
> I believe I am successfully executing the gpend command because my log file 
> contains the following:
> 
> pid =  12459
> gfile_path =  /tmp/gempak_params.12459
> About to run gdgrib
> GDFILE    Grid file                         
> /d1/dtcrt/data/gempak/caps/ssef_s4c0_nmm/ssef_s4c0_nmm_2010050600f030
> GFUNC     Scalar grid                       REFLCMP
> GDATTIM   Grid date/time                    f030
> GLEVEL    Grid level                        0
> GVCORD    Grid vertical coordinate          NONE
> GBTBLS    Input GRIB decoding tables        
> /home/dtcrt/svn/projects/HWT/HWT2010/ingestHome/params/ncepgrib129.tbl
> GBFILE    GRIB data file name               
> /d1/dtcrt/data/grib/caps/ssef_s4c0_nmm/native/20100506/g_000000/20100507_060000.grb
> VERCEN    PDS byte_4/byte_5/byte_6/byte_26  129/7/97/9
> PDSVAL    GRIB PDS grid identifier overrid  REFC%212
> PRECSN    Packing precision                 b/-5
> WMOHDR    WMO_ID/Origin_ID/DDHHMM
> CPYFIL    Grid file whose navigation is to
> PROJ      Map projection/angles/margins|dr
> GRDAREA   Area covered by grid
> KXKY      Number of grid points in x;y
> Parameters requested: GDFILE,GFUNC,GDATTIM,GLEVEL,GVCORD,GBTBLS,GBFILE,
> VERCEN,PDSVAL,PRECSN,WMOHDR,CPYFIL,PROJ,GRDAREA,KXKY.
> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> 
> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> 
> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> GEMPAK-GDGRIB> 
> GEMPAK-GDGRIB> [GDGRIB 4]  Warning:  grid not found in 
> grdnav.tbl...continuing.
> GRIB message written.
> Ran gdgrib
> About to run gpend
> Ran gpend
> 
> However, after the Python script ends, I see the following:
> 
> (taku) test 126 % ipcs
> 
> ------ Shared Memory Segments --------
> key        shmid      owner      perms      bytes      nattch     status
> 
> ------ Semaphore Arrays --------
> key        semid      owner      perms      nsems
> 
> ------ Message Queues --------
> key        msqid      owner      perms      used-bytes   messages
> 0x470a0004 0          rehak      666        0            0
> 0x440a0004 32769      rehak      666        0            0
> 0x00bc7bcf 83525634   dtcrt      666        0            0
> 0x02b90f66 83591171   dtcrt      666        0            0
> 0x00bc7c27 83656708   dtcrt      666        0            0
> 0x02b90f7c 83722245   dtcrt      666        0            0
> 
> (taku) test 127 % ps auxww | grep gplt
> dtcrt    12534  0.0  0.0  21896   792 pts/2    S    11:24   0:00 gplt
> dtcrt    12556  0.0  0.0  21896   784 pts/2    S    11:24   0:00 gplt
> dtcrt    19599  0.0  0.0   3568   696 pts/10   S+   11:35   0:00 grep gplt
> 
> Any ideas of what I might be missing?  This is the only GEMPAK processing I 
> am doing on this machine at this time.
> 
> Thanks!
> 
> 
> 


Ticket Details
===================
Ticket ID: MCL-633919
Department: Support GEMPAK
Priority: Normal
Status: Open