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

[IDDBrasil #XPZ-611749]: Re: 20060609: CONDUIT Latency (cont.)

Hi Waldenio,

> I�m very happy with these results !

Me too :-)

> I just did the 5-split changes.

OK.  After two cycles of ingest with a 5-way split, we
can see that the peak CONDUIT latencies were between 300
and 400 seconds.  Since this is not as good as what we would
like, I want to ask you to try the next test:

- replicate your 5-way split to idd.unidata.ucar.edu, and make
  the 5 new actions request from idd.cise-nsf.gov

After the change, your ~ldm/etc/ldmd.conf file will have 10
CONDUIT request lines: 5 will request from idd.cise-nsf.gov
and 5 will request from idd.unidata.ucar.edu.  Let's see if
this setup (which exactly matches that being used by Yoshihiro
Yamasaki at U. Averio) can reduce the largest latencies down
to under 60 seconds.  If it doesn't, the next test would be to
split the feed into sixths (1/6).

> Cheers,
> waldenio :-) :-) :-)
> --
> Waldenio Gambi de Almeida
> Instituto Nacional de Pesquisas Espaciais
> Centro de Previs�o de Tempo e Estudos Clim�ticos
> Divis�o de Operacoes
> Rodovia Pres. Dutra, km 40
> Cachoeira Paulista/SP - 12630-000
> Brasil
> Phone   : (12)3186-8541 / (12)9776-7805
> e-mail : address@hidden


Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
Unidata HomePage                       http://www.unidata.ucar.edu

Ticket Details
Ticket ID: XPZ-611749
Department: Support IDD Brasil
Priority: Normal
Status: Closed