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

20041124: efficient REQUEST entries



>From: Jamie Pelagatti <address@hidden>
>Organization: MIT
>Keywords: 200411241911.iAOJB37j026723 LDM ldmd.conf request

Hi Jamie,

>Im using LDM 6.1.0 and if I want to receive, say, two NEXRAD data streams, I
>could use either
>
>  request NEXRD2 "^L2.*(KIWX|KLOT)" 129.55.60.9
>  
>or 
>
>  request NEXRD2 "^L2.*KIWX" 129.55.60.9 
>  request NEXRD2 "^L2.*KLOT" 129.55.60.9 
>
>I'm curious which of these is more efficient or if it makes much difference.

Two comments are in order:

- there is little difference in splitting your request lines other than
  ending up with two rpc.ldmd process running.  If the volume of data
  you were asking for was significantly more (like in the CONDUIT stream
  of high resolution model output), my comment might be different.

- the regular expression you are using can/should be greatly simplified:

request NEXRAD2 "(KIWX|KLOT)"   129.55.60.9

(NB: NEXRAD2 is the default name for the feed as of LDM-6.1.0, but NEXRD2 and
CRAFT will work equally as well)

>I do observe that the latter causes two rpc.ldmd daemons to run rather than
>one. But it also occurs to me that two processes running asynchronously could
>perhaps be more efficient.
>Would you have any wisdom on this?

For the amount of data you are requesting, the effect of splitting the feed
request should be minimal.  This would depend on your network connection
and if there is any "packet shaping" (firewall throttling) being done
at your site (your machine, department, or campus, that is).

Cheers,

Tom Yoksas
--
NOTE: All email exchanges with Unidata User Support are recorded in the
Unidata inquiry tracking system and then made publicly available
through the web.  If you do not want to have your interactions made
available in this way, you must let us know in each email you send to us.