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

20031021: LEAD machine at IU



>From:  Jeff Weber <address@hidden>
>Organization:  UCAR/Unidata
>Keywords:  200310211621.h9LGLSkr013911

Jeff,

>LEAD Project machine " lead.extreme.indiana.edu " seems to be getting
>hammerred by the CONDUIT feed, should we split their feed into 5..?

Are you sure it is the CONDUIT feed causing this?  Compare the latency
plots for CONDUIT and, for instance, IDS|DDPLUS.  Both traces look the
same, so either the machine is ingesting everyting on on request line
(ala request ANY ".*" ...), or there is some network congestion occurring
that is affecting all feeds.

>Being at IU and an I2 hub I am surprised to see this type of
>latency..worse than to the UK !?

Yes, I am as well.  Do you know what their ldmd.conf request lines look
like?  Perhaps a little tuning would help.  Also, I see that they
are getting CONDUIT from atm.  It might be better to have them feed
from f5.aos.wisc.edu or flood.atmos.uiuc.edu.

Tom
--
+-----------------------------------------------------------------------------+
* Tom Yoksas                                             UCAR Unidata Program *
* (303) 497-8642 (last resort)                                  P.O. Box 3000 *
* address@hidden                                   Boulder, CO 80307 *
* Unidata WWW Service                             http://www.unidata.ucar.edu/*
+-----------------------------------------------------------------------------+