[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
Re: 20000323: 3 messages catted together....pqsurf
- Subject: Re: 20000323: 3 messages catted together....pqsurf
- Date: Thu, 23 Mar 2000 11:14:23 -0700 (MST)
On Thu, 23 Mar 2000, C. Vandersip wrote:
> Robb,
>
> Thanks for the advice. Things are still running smooth, 15 hours and
> counting. If it continues to chug along, I've got to believe it's related
> to remaking ldmd.pq, since when I recompiled ldm to add in gdbm support, I
> never recreated a new ldm queue, and maybe it needed that to safely
> interact with the pqsurf.pq. Or are they completely separate? Hmmm...
Chris,
The regular LDM queue is seperate from the pqsurf queue. Maybe only a
small section of the LDM queue was corrupt, that can happen.
Robb...
>
> Chris
>
> On Thu, 23 Mar 2000, Robb Kambic wrote:
>
> > Chris,
> >
> > There's a couple of possible answers here:
> >
> > - the pqsurf queue could of been corrupt, looks like you check that one.
> > - there is a strange product being sent to a decoder/file action in the
> > pqact for pqsurf. For some reason, it causes pqact to croak and then
> > pqsurf dies. I would make sure that pqact entries are as tight as
> > possible to try to eliminate these unexpected products. I don't know what
> > the names are of the products.
> >
> > - the MESSAGE_TYPE_UNKNOWN messages are usually caused by metar products
> > that start with FX, etc. Don't worry about these messages, they are not
> > the problem.
> >
> > Robb...
> >
> > On Thu, 23 Mar 2000, Unidata Support wrote:
> >
> > >
> > > ------- Forwarded Message
> > >
> > > >To: address@hidden
> > > >From: "C. Vandersip" <address@hidden>
> > > >Subject: Re: Problem with pqsurf signal 11 deaths (fwd)
> > > >Organization: .
> > > >Keywords: 200003231545.IAA02237
> > >
> > > Well, after vaporizing and recreating both ldmd.pq and pqsurf.pq and
> > > setting the pqsurf.pq size to the default value of 2MB, pqsurf's been
> > > running fine for about 12 hours now. Still getting the surface_split
> > > messages, but I'm guesssing they're a result of malformed metar reports
> > > within the feed. No signal 11 yet (fingers crossed).
> > >
> > > Chris
> > >
> > >
> > > On Thu, 23 Mar 2000, C. Vandersip wrote:
> > >
> > > > Add'l info:
> > > >
> > > > 1. Peter says he has not experienced this particular problem.
> > > >
> > > > 2. I'm getting numerous "surface_split: Can't handle
> > > > MESSAGE_TYPE_UNKNOWN"
> > > > messages in the pqsurf.log when pqsurf is running. Are these relevant
> > > > to
> > > > the problem? I just deleted both ldmd.pq and pqsurf.pq and remade them
> > > > to
> > > > see if that helps.
> > > >
> > > > Chris
> > > >
> > > > On Wed, 22 Mar 2000, C. Vandersip wrote:
> > > >
> > > > > Hi,
> > > > >
> > > > > I'm forwarding this message I sent to Peter Neilley regarding pqsurf
> > > > > deaths I've recently encountered after reconfiguring ldm-5.0.8 to
> > > > > create
> > > > > gdbm database metar files for the "weather" program. Any additional
> > > > > advice you might have would be much appreciated.
> > > > >
> > > > > Regards,
> > > > >
> > > > > Chris
> > > > >
> > > > > ###############################################################
> > > > > # Chris Vandersip #
> > > > > # Computer Research Specialist/Dept. Sysadmin #
> > > > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > > > # address@hidden (850)644-2522 #
> > > > > ###############################################################
> > > > >
> > > > > ---------- Forwarded message ----------
> > > > > Date: Wed, 22 Mar 2000 23:40:53 -0500 (EST)
> > > > > From: C. Vandersip <address@hidden>
> > > > > To: address@hidden
> > > > > Subject: Problem with pqsurf signal 11 deaths
> > > > >
> > > > > Peter,
> > > > >
> > > > > After running solidly for about 13 hours, pqsurf upped and died with a
> > > > >
> > > > > pqsurf[7919]: child 7921 terminated by signal 11
> > > > >
> > > > > message, the child being the pqact process forked from pqsurf. Have
> > > > > you
> > > > > encountered this before? Ever since, I cannot get pqsurf to run for
> > > > > more
> > > > > than an hour before dying again the same way. I've read in the
> > > > > ldm-support archives where this can be due to corrupt data in the feed
> > > > > that sometimes occurs on a feed-site basis, so I've switched my feed
> > > > > source to test that. Earlier, I increased my pqsurf.pq size from 2 to
> > > > > 20MB thinking that might help, but no luck there. Any insight would
> > > > > be
> > > > > greatly appreciated.
> > > > >
> > > > > Regards,
> > > > >
> > > > > Chris
> > > > >
> > > > > ###############################################################
> > > > > # Chris Vandersip #
> > > > > # Computer Research Specialist/Dept. Sysadmin #
> > > > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > > > # address@hidden (850)644-2522 #
> > > > > ###############################################################
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > > ###############################################################
> > > > # Chris Vandersip #
> > > > # Computer Research Specialist/Dept. Sysadmin #
> > > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > > # address@hidden (850)644-2522 #
> > > > ###############################################################
> > > >
> > > >
> > >
> > > ###############################################################
> > > # Chris Vandersip #
> > > # Computer Research Specialist/Dept. Sysadmin #
> > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > # address@hidden (850)644-2522 #
> > > ###############################################################
> > >
> > > >From address@hidden Wed Mar 22 21:49:27 2000
> > > Received: from lightning.met.fsu.edu (lightning.met.fsu.edu
> > > [128.186.5.17])
> > > by unidata.ucar.edu (8.8.8/8.8.8) with ESMTP id VAA08532
> > > for <address@hidden>; Wed, 22 Mar 2000 21:49:26 -0700 (MST)
> > > Organization: .
> > > Keywords: 200003230449.VAA08532
> > > Received: from localhost by lightning.met.fsu.edu (8.9.2/8.9.2) with
> > > ESMTP id XAA00891
> > > for <address@hidden>; Wed, 22 Mar 2000 23:48:57 -0500 (EST)
> > > X-Authentication-Warning: lightning.met.fsu.edu: cvander owned process
> > > doing -bs
> > > Date: Wed, 22 Mar 2000 23:48:57 -0500 (EST)
> > > From: "C. Vandersip" <address@hidden>
> > > To: address@hidden
> > > Subject: Problem with pqsurf signal 11 deaths (fwd)
> > > Message-ID: <address@hidden>
> > > MIME-Version: 1.0
> > > Content-Type: TEXT/PLAIN; charset=US-ASCII
> > >
> > > Hi,
> > >
> > > I'm forwarding this message I sent to Peter Neilley regarding pqsurf
> > > deaths I've recently encountered after reconfiguring ldm-5.0.8 to create
> > > gdbm database metar files for the "weather" program. Any additional
> > > advice you might have would be much appreciated.
> > >
> > > Regards,
> > >
> > > Chris
> > >
> > > ###############################################################
> > > # Chris Vandersip #
> > > # Computer Research Specialist/Dept. Sysadmin #
> > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > # address@hidden (850)644-2522 #
> > > ###############################################################
> > >
> > > ---------- Forwarded message ----------
> > > Date: Wed, 22 Mar 2000 23:40:53 -0500 (EST)
> > > From: C. Vandersip <address@hidden>
> > > To: address@hidden
> > > Subject: Problem with pqsurf signal 11 deaths
> > >
> > > Peter,
> > >
> > > After running solidly for about 13 hours, pqsurf upped and died with a
> > >
> > > pqsurf[7919]: child 7921 terminated by signal 11
> > >
> > > message, the child being the pqact process forked from pqsurf. Have you
> > > encountered this before? Ever since, I cannot get pqsurf to run for more
> > > than an hour before dying again the same way. I've read in the
> > > ldm-support archives where this can be due to corrupt data in the feed
> > > that sometimes occurs on a feed-site basis, so I've switched my feed
> > > source to test that. Earlier, I increased my pqsurf.pq size from 2 to
> > > 20MB thinking that might help, but no luck there. Any insight would be
> > > greatly appreciated.
> > >
> > > Regards,
> > >
> > > Chris
> > >
> > > ###############################################################
> > > # Chris Vandersip #
> > > # Computer Research Specialist/Dept. Sysadmin #
> > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > # address@hidden (850)644-2522 #
> > > ###############################################################
> > >
> > >
> > >
> > > >From address@hidden Wed Mar 22 22:45:25 2000
> > > Received: from lightning.met.fsu.edu (lightning.met.fsu.edu
> > > [128.186.5.17])
> > > by unidata.ucar.edu (8.8.8/8.8.8) with ESMTP id WAA14124
> > > for <address@hidden>; Wed, 22 Mar 2000 22:45:25 -0700 (MST)
> > > Organization: .
> > > Keywords: 200003230545.WAA14124
> > > Received: from localhost by lightning.met.fsu.edu (8.9.2/8.9.2) with
> > > ESMTP id AAA00948
> > > for <address@hidden>; Thu, 23 Mar 2000 00:44:55 -0500 (EST)
> > > X-Authentication-Warning: lightning.met.fsu.edu: cvander owned process
> > > doing -bs
> > > Date: Thu, 23 Mar 2000 00:44:55 -0500 (EST)
> > > From: "C. Vandersip" <address@hidden>
> > > To: address@hidden
> > > Subject: Re: Problem with pqsurf signal 11 deaths (fwd)
> > > In-Reply-To: <address@hidden>
> > > Message-ID: <address@hidden>
> > > MIME-Version: 1.0
> > > Content-Type: TEXT/PLAIN; charset=US-ASCII
> > >
> > > Add'l info:
> > >
> > > 1. Peter says he has not experienced this particular problem.
> > >
> > > 2. I'm getting numerous "surface_split: Can't handle MESSAGE_TYPE_UNKNOWN"
> > > messages in the pqsurf.log when pqsurf is running. Are these relevant to
> > > the problem? I just deleted both ldmd.pq and pqsurf.pq and remade them to
> > > see if that helps.
> > >
> > > Chris
> > >
> > > On Wed, 22 Mar 2000, C. Vandersip wrote:
> > >
> > > > Hi,
> > > >
> > > > I'm forwarding this message I sent to Peter Neilley regarding pqsurf
> > > > deaths I've recently encountered after reconfiguring ldm-5.0.8 to create
> > > > gdbm database metar files for the "weather" program. Any additional
> > > > advice you might have would be much appreciated.
> > > >
> > > > Regards,
> > > >
> > > > Chris
> > > >
> > > > ###############################################################
> > > > # Chris Vandersip #
> > > > # Computer Research Specialist/Dept. Sysadmin #
> > > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > > # address@hidden (850)644-2522 #
> > > > ###############################################################
> > > >
> > > > ---------- Forwarded message ----------
> > > > Date: Wed, 22 Mar 2000 23:40:53 -0500 (EST)
> > > > From: C. Vandersip <address@hidden>
> > > > To: address@hidden
> > > > Subject: Problem with pqsurf signal 11 deaths
> > > >
> > > > Peter,
> > > >
> > > > After running solidly for about 13 hours, pqsurf upped and died with a
> > > >
> > > > pqsurf[7919]: child 7921 terminated by signal 11
> > > >
> > > > message, the child being the pqact process forked from pqsurf. Have you
> > > > encountered this before? Ever since, I cannot get pqsurf to run for more
> > > > than an hour before dying again the same way. I've read in the
> > > > ldm-support archives where this can be due to corrupt data in the feed
> > > > that sometimes occurs on a feed-site basis, so I've switched my feed
> > > > source to test that. Earlier, I increased my pqsurf.pq size from 2 to
> > > > 20MB thinking that might help, but no luck there. Any insight would be
> > > > greatly appreciated.
> > > >
> > > > Regards,
> > > >
> > > > Chris
> > > >
> > > > ###############################################################
> > > > # Chris Vandersip #
> > > > # Computer Research Specialist/Dept. Sysadmin #
> > > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > > # address@hidden (850)644-2522 #
> > > > ###############################################################
> > > >
> > > >
> > > >
> > > >
> > >
> > > ###############################################################
> > > # Chris Vandersip #
> > > # Computer Research Specialist/Dept. Sysadmin #
> > > # Rm. 024, Dept. of Meteorology, Florida State University #
> > > # address@hidden (850)644-2522 #
> > > ###############################################################
> > >
> > >
> > > ------- End of Forwarded Message
> > >
> >
> > ===============================================================================
> > Robb Kambic Unidata Program Center
> > Software Engineer III Univ. Corp for Atmospheric
> > Research
> > address@hidden WWW: http://www.unidata.ucar.edu/
> > ===============================================================================
> >
>
> ###############################################################
> # Chris Vandersip #
> # Computer Research Specialist/Dept. Sysadmin #
> # Rm. 024, Dept. of Meteorology, Florida State University #
> # address@hidden (850)644-2522 #
> ###############################################################
>
===============================================================================
Robb Kambic Unidata Program Center
Software Engineer III Univ. Corp for Atmospheric Research
address@hidden WWW: http://www.unidata.ucar.edu/
===============================================================================
