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

[#RFZ-122547]: Re: Re: More data to send from CMA

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: [#RFZ-122547]: Re: Re: More data to send from CMA
  • Date: Sat, 07 Apr 2007 15:57:55 -0600
Doug,

> I'll set ultrazone's ldm to stop
> requesting babj from ECMWF, and instead request from CMA's ldm.

Is CMA running the "balance" program to connect port 8080 to port
388 and will your REQUEST entries connect to port 8080?  If not,
then it's likely that the throughput will be insufficient.  The
experiment that Mike Schmidt did (having CMA's LDM connect to
port 8080 at NCAR) demonstrated that traffic using port 388 was
being restricted.  It seems to me that this condition is likely
to still exist.

> ECMWF is acting as a surrogate, inserting data into its ldm from UKMO
> and JMA in addition to its own data.  ECMWF ftp's data from JMA, then
> inserts it into ldm (this is what they were doing for CMA).  Since UKMO
> shares computational resources on site, ECMWF has agreed to insert
> UKMO's data into its ldm server, to avoid spinning up 2 servers at
> the same physical site.

Interesting.  I don't see anything wrong with two LDM servers at
the same site, however.  If I understand correctly, the data 
supplied by each would be disjoint from the other -- consequently,
the total bandwidth usage would be the same as one LDM server
supplying everything.

Is spinning up an LDM server all that difficult?  If so, then I
have more work to do.  :-)

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: RFZ-122547
Department: Support IDD TIGGE
Priority: Normal
Status: Open