[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Datastream #OKF-212399]: core dumps on unidata2-new



I spoke too soon.  The DCGRIB -11 errors are occurring again and I'm looking at 
it now.

-Michael




> Thanks!
> 
> Douglas
> 
> On 8/19/13 4:55 PM, Unidata GEMPAK Support wrote:
> >  From the /var/adm/messages line
> >
> > Aug 19 18:03:14 unidata2-new.ssec.wisc.edu genunix: [ID 603404 kern.notice] 
> > NOTICE: core_log: dcgrib2[20848] core dumped: 
> > /data/core/core.dcgrib2.20848.u8858
> >
> > I trace process ID 20848 to the following entries in ~ldm/logs/ldmd.log
> >
> > Aug 19 18:03:14 unidata2-new.ssec.wisc.edu pqact[18665] ERROR: pipe_put(): 
> > write error: pid=20848, cmd=(decoders/dcgrib2 -v 1 -d 
> > data/gempak/logs/dcgrib2_fnmoc.log -e 
> > GEMTBL=/home/gempak/NAWIPS/gempak/tables)
> > Aug 19 18:03:14 unidata2-new.ssec.wisc.edu pqact[18665] ERROR: Deleting 
> > failed PIPE entry: pid=20848, cmd="decoders/dcgrib2 -v 1 -d 
> > data/gempak/logs/dcgrib2_fnmoc.log -e 
> > GEMTBL=/home/gempak/NAWIPS/gempak/tables" [filel.c:288]
> > Aug 19 18:03:14 unidata2-new.ssec.wisc.edu pqact[18665] WARN: child 20848 
> > terminated by signal 11
> >
> > and in ~ldm/data/gempak/logs/ I see the dcgrib2_fnmoc.log file was over 2 
> > GB in size!  I created a new logfile for the fnmoc feed and so far have not 
> > see any further DCGRIB -11 errors in the new log.  I think this problem was 
> > caused by not rotating the GEMPAK grib decoder logs.
> >
> > -Michael James
> > Unidata
> >
> >
> >
> >
> >
> >> Hi Tom
> >>
> >> Thanks for all the information. I'll make some notes.
> >>
> >> Be well.
> >>
> >> Douglas
> >>
> >> On 8/19/13 12:26 PM, Unidata GEMPAK Support wrote:
> >>> Hi Douglas,
> >>>
> >>> re:
> >>>> We are seeing a lot of core dump messages this morning on unidata2-new
> >>>> for dcgrib2 procs.
> >>> OK, thanks for letting us know.  Since 'dcgrib2' is a GEMPAK decoder, I
> >>> will pass the information along to our GEMPAK person.
> >>>
> >>> re:
> >>>> I thought I might try to restart xcd via xcdadmin but I do not know how
> >>>> xcd is started on unidata2-new.
> >>> Why did you want to restart XCD?  Perhaps you thought that 'dcgrib2' is
> >>> not a McIDAS XCD routine?
> >>>
> >>> re:
> >>> It appears to be running under ldm but
> >>>> the config file shows oper as the admin user:
> >>>>
> >>>> xcd_admin="oper"
> >>>> xcd_admin_home="/home/oper"
> >>>> xcd_admin_profile="$xcd_admin_home/.profile"
> >>>>
> >>>> It does not look like the file was updated recently.
> >>>>
> >>>> Here is a sample of the error messages we are seeing in 
> >>>> /var/adm/messages:
> >>>>
> >>>> Aug 19 16:52:30 unidata2-new.ssec.wisc.edu genunix: [ID 603404 
> >>>> kern.notice] NOTICE: core_log: dcgrib2[6390] core dumped: 
> >>>> /data/core/core.dcgrib2.6390.u8858
> >>>> Aug 19 16:52:36 unidata2-new.ssec.wisc.edu genunix: [ID 603404 
> >>>> kern.notice] NOTICE: core_log: dcgrib2[6656] core dumped: 
> >>>> /data/core/core.dcgrib2.6656.u8858
> >>>>
> >>>> The error state has existed for almost 4 hours as I type this.
> >>> The McIDAS-XCD decoders running on unidata2-new are from the Unidata
> >>> McIDAS-X/XCD distribution.  This installation eliminates the use of
> >>> an 'oper' account; the running of the XCD data monitors and decoding
> >>> processes is handled in the 'ldm' account, and the configuration of
> >>> the XCD data monitors, etc. is handled in the 'mcidas' account.  I
> >>> eliminated the 'oper' part of McIDAS installations in the Unidata
> >>> McIDAS-X/XCD release as soon as XCD was allowed (by Bob Fox and
> >>> then Tom Achtor) to be bundled with Unidata McIDAS.  This was over
> >>> 12 years ago (I think).
> >>>
> >>> Anyway, the problem is with a GEMPAK decoder run from an 'ldm'
> >>> pattern-action file.  I will let our GEMPAK guy know that the
> >>> installation on the machine needs attending to.
> >>>
> >>> re:
> >>>> Thanks in advance.
> >>> Sorry for the confusion...
> >>>
> >>> Cheers,
> >>>
> >>> Tom
> >>> --
> >>> ****************************************************************************
> >>> Unidata User Support                                    UCAR Unidata 
> >>> Program
> >>> (303) 497-8642                                                 P.O. Box 
> >>> 3000
> >>> address@hidden                                   Boulder, CO 80307
> >>> ----------------------------------------------------------------------------
> >>> Unidata HomePage                       http://www.unidata.ucar.edu
> >>> ****************************************************************************
> >>>
> >>>
> >>> Ticket Details
> >>> ===================
> >>> Ticket ID: OKF-212399
> >>> Department: Support GEMPAK
> >>> Priority: Normal
> >>> Status: Open
> >>>
> >>
> >> --
> >>
> >> SSEC Data Center            Email:address@hidden
> >> UW Madison                  Web:http://www.ssec.wisc.edu/datacenter
> >> 1225 W. Dayton              Phone:  (608) 262-0502
> >> Madison, Wisconsin 53706    Fax:    (608) 263-6738
> >>
> >>
> > Ticket Details
> > ===================
> > Ticket ID: OKF-212399
> > Department: Support GEMPAK
> > Priority: Normal
> > Status: Open
> >
> 
> 
> --
> 
> SSEC Data Center            Email:address@hidden
> UW Madison                  Web:http://www.ssec.wisc.edu/datacenter
> 1225 W. Dayton              Phone:  (608) 262-0502
> Madison, Wisconsin 53706    Fax:    (608) 263-6738
> 
> 


Ticket Details
===================
Ticket ID: OKF-212399
Department: Support GEMPAK
Priority: Normal
Status: Open