On Wed, 2009-10-07 at 18:27 -0400, Michael Dross wrote:
> I am running Unidata NOAAPORT v1.3.0 along with LDM. I suspect most people
> using the Unidata NOAAPORT code are running 1.3.0, since it was the latest
> version and
> was released in 2007.
>
> What happens is that readnoaaport shuts down and this cascades and shuts
> down the whole LDM system.
>
> >From my log file from the most recent shutdown.
>
> Oct 07 22:16:54 noaaport2 rpc.ldmd[5549] NOTE: child 5557 terminated by
> signal 11: readnoaaport -m 224.0.1.3
> Oct 07 22:16:54 noaaport2 rpc.ldmd[5549] NOTE: Killing (SIGTERM) process
> group
> Oct 07 22:16:54 noaaport2 rtstats[5559] NOTE: Exiting
> Oct 07 22:16:54 noaaport2 rpc.ldmd[5549] NOTE: Exiting
> Oct 07 22:16:54 noaaport2 rpc.ldmd[5549] NOTE: Terminating process group
>
>
>
> Maybe a termination signal 11 from the readnoaaport exe might provide a
> clue?
>
Since I don't have access to the readnoaaport code, I can't do much
here. But a signal 11 is a segmentation fault and its occuring on the
NWSTG2 channel. I know NCEP has been doing some things with the CCB
header and multiple routing sequences. I know in the past most software
only accounted for one or at most two. If the readnoaaport code only
accounts for a limited number, this could be a problem. I can't think
of much else that would cause a problem. I can trap for this here and
see if this is an issue.
________________________________________________________________________
Daniel Vietor
Mail: devo@xxxxxxxxxxxxx
Unisys Corp
Title: Engineer/Meteorologist
2476 Swedesford Rd
Phone: 610-648-3623
Malvern PA 19355
Fax: 610-695-5524
