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

Re: 20010623: LDM 5.1.4 problem?

Gilbert Sebenste wrote:
> 
> Hi Anne,
> 
> > Hi Gilbert,
> >
> > Now I know what happened.  I did make a change to pqing in 5.1.3 when an
> > NWS person couldn't get his LDM to handle the AFOS products properly.
> > There was an error in the code so that AFOS products were instead being
> > treated as WMO products.  I guess you want the old behavior.
> 
> Yep!
> 
> > But, I think this means that you could still have pqing handle the
> > current products the old way.  Although this is a little perverted, if
> > you invoked pqing with the -f NMC2 or -f NMC3 feedtype, even though it's
> > not really one of those feedtypes, I believe you will get the old
> > behavior.  That is, the line in ldmd.conf would be:
> >
> > exec    "pqing -f NMC2 -v -b 9600 -p none /dev/ttyb"
> 
> And here's what happened:
> 
> -------------
> That seems to do the trick!   5.1.4 is running now on the main ingest
> machine.
> 
> Now the big question, do you have a copy of the NWWS announcement about
> PIL's you could e-mail me?  I accidently deleted it earlier this month!
> 
> Thanks-
> 
HI Gilbert,

I'm glad that worked, but, um, well, let me back pedal a little bit.  I
should have thought this through further.

Caveat #1: If products are being ingested under a 'false' feedtype, this
will cause problems if you're propagating them.  The NMC2 name is now
being used for the CONDUIT data, which is not currently being
propagated.  But when that occurs there will be confusion.  

Caveat #2: If you and/or your upstream site are sending statistics (and
we hope that you are), this will skew the statistics.  Is there a lot of
data on this feed?  If it's small it's no big deal, but if it's large,
it could be a problem for us.

Generally, sending stuff under false feedtypes like this is not
recommended (even though I just did that with you!)  An alternative is
for your upstream site to make what is a very simple change to pqing. 
Obviously they would need to rebuild the software, and remember to make
the change every time they upgrade.  If they would like to do this, let
me know and I'll give them the change.

So, I realize now that I must urge you not to propagate these products.

Regarding the notice, I don't have that any more.  But we've got
archives of all that junk.  See
http://www.unidata.ucar.edu/glimpsedocs/glimpse.html.  From there,
search the nws-changes archive.  Let me know if you can't find it.

Anne
-- 
***************************************************
Anne Wilson                     UCAR Unidata Program            
address@hidden                 P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************