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

Re: 19990716: CONDUIT latencies (fwd)

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: 19990716: CONDUIT latencies (fwd)
  • Date: Tue, 20 Jul 1999 07:52:42 -0600 (MDT)

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

---------- Forwarded message ----------
Date: Tue, 20 Jul 1999 12:41:26 GMT
From: "David J. Knight" <address@hidden>
To: address@hidden, address@hidden
Subject: Re: 19990716: CONDUIT latencies


Hi Chiz,
     Splitting up the feed into two requests helped a little.
Changing to getting the feed from nasa helped a little more.
But, as I'm sure you are aware, latencies still get to 3600 sec
and data gets dropped. I assume this is a result of the limited
bandwidth of the T1 line between NCEP and NASA. In order to take
full advantage of this feed, it needs to be more reliable
and drop less data. If the T1 line cannot be upgraded,
a couple of other possibilities occur to me -
which you may have already considered.
1) we could simply increase the time-to-live in the
product queue. This would require larger product queues
at each location receiving the feed, but might be worth it.
One problem might be that this would also increase the
time-to-live of all other products in the queue. While
a wait for model data is acceptable, perhaps a wait for
realtime observations might not be to all sites.
2) Perhaps pqinsert at ncep could be slowed down. Instead
of dumping the entire mrf run into the product queue when
it arrives, perhaps it could be inserted more slowly over
a period of time. Simply sleeping for a few seconds between
insertion of each product might work.

Anyway, just some random thoughts. Thanks for your help,
the feed is more consistant now.

David

David Knight
Department of Earth and Atmospheric Sciences   Tel: (518)-442-4204
SUNYA   ES-228                                 Fax: (518)-442-4494
Albany, NY  12222                              Email: address@hidden