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

[Fwd: Re: 20000712: Wierd ldm behavior] (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: Thu, 13 Jul 2000 08:45:02 -0600
From: Anne Wilson <address@hidden>
To: address@hidden
Subject: [Fwd: Re: 20000712: Wierd ldm behavior]

Hi Robb,

Here's Tom's response to the issue we discussed
yesterday afternoon.  Do you have any ideas or
comments?

Anne

-------- Original Message --------
Subject: Re: 20000712: Wierd ldm behavior
Date: Thu, 13 Jul 2000 08:01:24 -0400 (EDT)
From: Tom McDermott
<address@hidden>
Organization: UCAR/Unidata
To: Anne Wilson <address@hidden>
CC: address@hidden

On Wed, 12 Jul 2000, Anne Wilson wrote:

> This sounds odd.  The changes made between 5.0.9 and 5.0.10 were minimal, and
> none applied to the portions of the code to which these messages pertain.  
> There
> are usually a few (not thousands!) of RECLASS messages upon start up of an
> upstream or downstream site while the sites renegotiate what products to send.
> Otherwise, by themselves, the RECLASS and pq_sequence error messages usually
> indicate  network problems.
> 
> Can you duplicate this problem?  I.e., if you switch back to 5.0.10 does it
> happen again?   I'm wondering if now that gissun is caught up these problems
> won't appear again...

Anne,

Yes, as a matter of fact; the full story and why I
was fairly certain it
was related to v. 5.0.10:  I first atttempted to
upgrade at the end of
April from v. 5.0.8 to 5.0.10 and observed the
same phenomenon.  At the
time I thought it may have been a fluke or was
related somehow to network
congestion at the end of the semester.  So I went
back to 5.0.8 and
problem vanishes.  Then Colgate disconnected from
our ldm server at the
end of the semester (approx beginning of May). 
Then I reattempted the
install of 5.0.10 at end of May.  As I said, no
problems for 6 weeks until
Colgate reconnects on Monday.  Then, bam, the same
thing happened.  So I
don't think it's necessary for me to repeat the
experiment a 3rd time.  I
know what will happen.

It was my understanding as well that the changes
between the most recent
ldm versions were fairly minor.  Perhaps there is
something peculiar about
gissun, but there haven't been any major problems
w/ gissun this week
under 5.0.9, RECLASSING occasionally only when the
models come in (and
that could be due mainly to upstream delays).  An
interesting test would
be for Adam to upgrade to 5.0.10 and see what
happens.

Tom