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

[LDM #KIO-429794]: LDM configuration

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: [LDM #KIO-429794]: LDM configuration
  • Date: Thu, 18 Oct 2007 14:41:01 -0600
Niles,

> I have a question about ldm configuration.
> 
> Say we are ingesting level two nexrad radar data for the
> Washington DC radar, KLWX.
> 
> What we currently do is to have our request include all
> radars in ldmd.conf :
> 
> request NEXRD2  ".*" rasp.rap.ucar.edu
> 
> And then only process KLWX in the pqact.conf :
> 
> NEXRD2  
> ^L2-(.*)/(KLWX)/([0-9][0-9][0-9][0-9][0-1][0-9][0-3][0-9])([0-2][0-9][0-5][0-9][0-5][0-9])/(.*)/(.*)
> FILE    /home/ldm/data/radar/\2/\3/\4_\6.\1
> 
> We're realizing that what that probably does is make for a much bigger 
> product queue,
> which we'd like to avoid. So instead of doing that, we'd like to request only 
> that radar
> in the ldmd.conf file :
> 
> request NEXRD2  "KLWX" rasp.rap.ucar.edu
> 
> and then leave the pqact.conf file open to any incoming radars :
> 
> NEXRD2  
> ^L2-(.*)/(.*)/([0-9][0-9][0-9][0-9][0-1][0-9][0-3][0-9])([0-2][0-9][0-5][0-9][0-5][0-9])/(.*)/(.*)
> FILE    /home/ldm/data/radar/\2/\3/\4_\6.\1
> 
> That way if we do add another radar we only have to edit ldmd.conf and not 
> pqact.conf.
> 
> My question is - is this a good thing to do,

Requesting only the data that you want will reduce bandwidth
demand, upstream CPU usage, downstream CPU usage, and product
-queue size -- all very good things!

> will it have any effects
> that we have not thought of.

You will not be able to relay any data products that you have
not requested.

> I'm guessing load on the upstream host, rasp, might go
> up slightly because it has to do more sorting,

I think the CPU usage will actually be less.

> but that network traffic would drop and
> product queue size would drop significantly (the desired result for us).

Without question.

> So - is this a good move to make?

Absolutely!  Providing you're not interested in relaying the
data that you're not receiving.

> Niles.
> 
> --
> 
> Work (USA) 303 497 2808    email address@hidden
> Home (USA) 720 938 7332    email address@hidden (simple short text, receive 
> only)


Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: KIO-429794
Department: Support LDM
Priority: Normal
Status: Closed