Re: Linux LDM (fwd)

===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================

---------- Forwarded message ----------
Date: Mon, 30 Aug 1999 14:27:30 -0500 (CDT)
From: Gilbert Sebenste <address@hidden>
To: Robb Kambic <address@hidden>
Subject: Re: Linux LDM (fwd)

Robb,

> Gilbert,
> 
> Are you seeing any strange behavior on your Linux boxes with the ldm-5.0.8
> release?  Is the queue growing, using more disk space, etc.  What version
> of Linux are you using, size of queue, amount of ram, etc?
> 
> Thanks,
> Robb...
> 
> ===============================================================================
> Robb Kambic                              Unidata Program Center
> Software Engineer III                    Univ. Corp for Atmospheric Research
> address@hidden                   WWW: http://www.unidata.ucar.edu/
> ===============================================================================
> 
> ---------- Forwarded message ----------
> Date: Fri, 27 Aug 1999 17:01:50 -0500 (CDT)
> From: David Wojtowicz <address@hidden>
> 
> > David,
> > 
> > Are you still running the LDM with mmap files so _MAPRGNS was
> > defined for  Linux?  I have a site with a 650megabyte queue that
> > having a problem. The queue seems to grow even though NOGROW was set.
> > Eventually the machine runs out of disk space.  I'm trying to determine if
> > the problem is the site's configuration or it's the _MAPRGNS flag being
> > set?  What's your experiences?
> > 
> 
> Hi Robb,
> 
> No, I'm not seeing any queue growth on Linux with _MAPRGNS set.
> The actual size of the queue file is slightly larger than is
> specified in ldmadmin, but that seems to be a rounding off kind
> of thing.  They don't seem to be growing during operation. 
> I checked the queue size on the NMC2 ingest machine which
> has been running for over a month without a ldm restart and
> then remade the queue.  It is the same size after remaking
> as before, so it had not grown any during a long period of
> operation.  That is a RH6.0 machine with a ~800MB queue.
> 
> I also checked our main ldm machine running 5.3 with a ~300MB
> queue and the same is true there, though it gets restarted
> more frequently.   So, I'd say, no, we're not seeing any
> queue growth.
> 
> Note...I'm still using a pre-release version of 5.0.8 with my own mods to
> define _MAPRGNS...had not gotten around to installing the final one yet so
> there may be a difference there.  You might ask Gilbert Sebenste.
> I helped him install the new 5.0.8 on both of his Linux boxes last
> week.
> 
> --david

I am using Red Hat Linux 6.0 on both my boxes, latest patches/kernel
update with the official 5.0.8 update (thanks, David...I can now install
it by myself!). I have a 275 MB queue on both machines.. I have 250 MB of
RAM on both machines; one is a single PII 450, the other dual PII 450
MHZ processors machine from SW Technologies. I just noticed that my ldm.pq
is 281 MB, so it is exceeding the limit I have specified slightly. But I
haven't had any problems with it that I am aware of. 

Gilbert

*******************************************************************************
Gilbert Sebenste                                                     ********
Internet: address@hidden    (My opinions only!)                     ******
Staff Meteorologist, Northern Illinois University                      ****
Work phone: 815-753-5492                                                ***
*******************************************************************************